Characterization and Calibration Setup for a Medical Laser System

(1)

CHARACTERIZATION AND CALIBRATION SETUP FOR A MEDICAL LASER SYSTEM

Faculty of Information Technology and Communication Sciences Master of Science Thesis May 2020

(2)

ABSTRACT

Matius Hurskainen: Characterization and Calibration Setup for a Medical Laser System Master of Science Thesis

Tampere University Electrical Engineering May 2020

Laser devices and systems have become increasingly important in industrial, medical and consumer electronic applications. In the near future, the demand of solutions such as facial recognition and automotive rangefinders will increase. Research in microprocessor technology and quantum computing on the other hand suggests of entirely new kind of requirements in the coming years.

One of the most rapidly growing branches of laser technology are medical devices. Applications in this field include, among many others, tattoo removal, brain surgery, diagnosis of disease, ophthalmic instruments and cancer treatment.

The tasks of this work is the development of methods and equipment for characterization and calibration of an ophthalmic, medical laser system. The aim is to both enable safe and reliable operation of this product, as well as to allow for seamless replacement of either of the system’s key components.

Such a setup has not yet existed, and developing one is crucial for enabling effective and repeatable manufacturing of the product.

In this work, laser fundamentals and electrical characteristics are explored and regulatory challenges posed by the intended use of medical purposes are discussed. Further properties related to laser systems and actual driving of laser diodes, as well as the need for accompanying electronics are in turn analyzed.

The structure of the actual laser system and necessary characterization and calibration processes are identified. This includes the requirements for the accuracies of said processes, in order to make sure the target parameters for the laser system are met, even if either of the system’s components is replaced.

This functionality is found to require characterization of the laser system’s measurement electronics.

To validate the chosen methods for characterizing these components, a test apparatus is designed and built. Initial testing with this appliance and other laboratory equipment indicates the suitability of the chosen methods and electronic solutions for the purpose.

Based on these promising results, an automated characterization setup is designed and built to replace the test apparatus. This new setup is based around an embedded computer, with several additional components and dedicated software. The solution is then used for final testing of two laser systems.

Due to incomplete product firmware, verification of photodiode current measurement was not possible. However, the targets for laser irradiances, as well as spot size measurements were all achieved.

Keywords: laser system, characterization, calibration, medical device, irradiance

The originality of this thesis has been checked using the Turnitin OriginalityCheck service.

(3)

TIIVISTELMÄ

Matius Hurskainen: Karakterisointi- ja kalibrointilaitteisto lääketieteelliselle laserjärjestelmälle Diplomityö

Tampereen yliopisto Sähkötekniikka Toukokuu 2020

Laserlaitteista ja järjestelmistä on tullut yhä tärkeämpiä teollisuudessa, lääketieteessä ja kulutuselektroniikassa. Erityisesti sovelluskohteiden, kuten kasvontunnistuksen ja autoteollisuuden etäisyysmittarien, kysyntä tulee kasvamaan lähitulevaisuudessa. Mikroprosessoritekniikan ja kvanttitietokoneiden tutkimus puolestaan viittaa kokonaan erityyppisiin vaatimuksiin tulevina vuosina.

Yksi nopeimmin kasvavista lasertekniikan osa-alueista ovat lääketieteelliset laitteet.

Sovelluskohteisiin tällä alalla kuuluvat, monien muiden lisäksi, tatuointien poisto, aivokirurgia, sairauksien diagnosointi, oftalmiset laitteet ja syöpähoito.

Tehtävänä tässä työssä on kehittää menetelmät ja laitteisto oftalmisen, lääketieteellisen laserjärjestelmän karakterisoimiseksi ja kalibroimiseksi. Tavoitteena on mahdollistaa sekä tuotteen turvallinen ja luotettava käyttö, että järjestelmän kumman tahansa pääkomponentin vaihtaminen saumattomasti. Tällaista laitteistoa ei ole vielä ollut olemassa, ja sen kehittäminen on ensiarvoisen tärkeää laserjärjestelmän luotettavan ja toistettavan valmistamisen mahdollistamiseksi.

Työssä tutustutaan lasertekniikan perusteisiin ja lasereiden sähköisiin ominaisuuksiin. Sääntelyn aiheuttamia, aiotusta lääketieteellisestä käyttötarkoituksesta johtuvia haasteita pohditaan.

Laserjärjestelmiin ja laserdiodien varsinaiseen ajamiseen liittyviä erityisiä ominaisuuksia sekä tarvetta oheiselektroniikalle puolestaan analysoidaan.

Laserjärjestelmän rakenne ja tarvittavat karakterisointi- ja kalibrointiprosessit tunnistetaan. Tämä sisältää vaatimukset prosessien tarkkuuksille, jotta voidaan varmistua laserjärjestelmän tavoiteparametrien saaavuttamisesta, myös vaihdettaessa kumpi tahansa järjestelmän pääkomponentti toiseen.

Tämän toiminnallisuuden havaitaan vaativan laserjärjestelmän mittauselektroniikan karakterisointia. Näiden komponenttien karakterisointiprosessien validoimiseksi suunnitellaan ja rakennetaan testilaite. Tällä ja muilla laboratoriolaitteilla suoritetut ensitestit osoittavat valittujen menetelmien ja elektroniikkaratkaisujen soveltuvan tarkoitukseen.

Näiden lupaavien tulosten pohjalta suunnitellaan automaattinen karakterisointilaitteisto testilaitteen korvaajaksi. Tämä uusi laitteisto pohjautuu sulautettuun tietokoneeseen useiden lisäkomponenttien ja tarkoituksenmukaisen ohjelmiston tukemana. Laitteiston avulla suoritetaan lopullinen testaus kahdelle laserjärjestelmälle.

Tuotteen laiteohjelmiston keskeneräisyydestä johtuen valodiodin virranmittausta ei kyetty verifioimaan. Sen sijaan kaikki tavoitteet laserien irradiansseille ja spottikoon mittaukselle täyttyivät.

Avainsanat: laserjärjestelmä, karakterisointi, kalibrointi, lääketieteellinen laite, irradianssi Tämän julkaisun alkuperäisyys on tarkastettu Turnitin OriginalityCheck -ohjelmalla.

(4)

PREFACE

I am grateful for Seppo Orsila and Dr. Petteri Uusimaa for giving me the opportunity to work on the subject of this thesis. I would like to thank my examiners Prof. Karri Palovuori and Univ.

Lect. Erja Sipilä for their invaluable feedback and flexibility during the writing of this thesis.

For their expertise, directions and abundant guidance, I wish to thank my supervisors Visa Kaivosoja and Jukka-Pekka Alanko.

I’m thankful to Tero Vierimaa, Riina Hietikko and Esa Juntunen for their support in, among many things, software development and proofreading of this thesis. Thank you also to Dr. Jari Nikkinen and Antti Saarela for taking the time to discuss semiconductor lasers with me.

Lastly, I wish to thank all my friends and family for supporting me throughout my studies and making life that much greater. Most of all, thank you Heidi for always being there.

At Tampere, 19th May 2020

Matius Hurskainen

(5)

LIST OF FIGURES

2.1 Stimulated emission . . . 4

2.2 Four state laser operating principle . . . 5

2.3 Example semiconductor laser chip structure . . . 6

2.4 Semiconductor laser pn-junction at thermal equilibrium . . . 7

2.5 Semiconductor laser module (Courtesy of COMPANY) . . . 7

2.6 Laser module pinout . . . 8

2.7 Example laser module LIV graph . . . 9

2.8 LIV graphs of a laser module in different temperatures . . . 10

2.9 Output spectrums of a laser module in different temperatures . . . 10

4.1 LIV-graphs of laser modules produced in the same batch (Courtesy of COMPANY) . . . 16

4.2 Lifetime test of a laser chip showing aging (Courtesy of COMPANY) . . . 16

4.3 Block diagram of a laser system . . . 17

4.4 PD current measurement circuit . . . 19

4.5 Faster PD measurement circuit . . . 20

4.6 Spot size measurement circuit . . . 21

4.7 Graphical representation of a LD power calibration table . . . 22

5.1 System description . . . 24

5.2 Simplified LDD circuitry . . . 25

5.3 Actual spot size measurement circuit . . . 26

5.4 Actual photodiode current measurement circuit . . . 26

5.5 Processes responsible of TB irradiance . . . 29

6.1 TB transmission characterization results . . . 33

6.2 AB transmission characterization results . . . 33

6.3 TB power validation . . . 34

6.4 AB power validation . . . 34

6.5 Spot size characterization circuit . . . 36

6.6 Spot size measurement circuit characterization test . . . 37

6.7 Spot size validation results . . . 39

6.8 Spot size validation results with device 1 and BSU 2 . . . 40

6.11 Spot size validation in percentages with device 1 and BSU 1 . . . 41

(8)

6.15 Photodiode current characterization circuit . . . 43

6.16 Photodiode current measurement circuit characterization test . . . 44

7.1 Spot size verification of device 1 and BSU 1 with different beam profiler camera 47 7.2 Spot size verification of device 1 and BSU 2 with different beam profiler camera 47 7.3 Spot size verification of device 2 and BSU 2 with different beam profiler camera 48 7.4 Spot size verification of device 2 and BSU 1 with different beam profiler camera 48 7.5 Captures of a 7 mm and a 1.5 mm spot . . . 49

7.6 Spot size difference verification . . . 50

7.7 TB power verification of device 1 and BSU 1 with different power meter . . . 50

7.11 AB power verification of device 1 and BSU 1 with different power meter . . . 52

7.15 TB irradiance verification . . . 54

(9)

LIST OF SYMBOLS AND ABBREVIATIONS

AB Aiming beam

ADC Analog-to-digital converter BSU Beam shaping unit

CB Certification body CC Constant current CD Compact disk

CE French Conformité Européenne, European Conformity CP Constant power

CV Constant voltage

CVC Current to voltage converter CW Continuous wave

DAC Digital-to-analog converter EEA European Economic Area EU European Union

FDA U.S. Food and drug administration GaAs Gallium arsenide

IC Integrated circuit

IEC International Electrotechnical Commission

IECEE IEC System for Conformity Assessment Schemes for Electrotechnical Equipment and Components

IR Infrared

LASER Light amplification by stimulated emission of radiation LD Laser diode

LDD Laser diode driver LED Light emitting diode

LiDAR Light detection and ranging

LIV Light radiant power, current, voltage LSB Least significant bit

MCM Multi chip module

(10)

MCU Microcontroller unit NB Notified body NIR Near-infrared PCB Printed circuit board

PD Photodiode

PMA Premarket approval PMN Premarket notification SPI Serial peripheral interface TB Treatment beam

TEC Thermoelectric cooler UI User interface

(11)

1 INTRODUCTION

Ever since the development of the first semiconductor lasers in 1962, this field has grown to encompass an ever expanding variety of technologies from daily life to telecommunications, industrial and military applications. The success that began in consumer electronics from compact disk (CD) players and barcode scanners has evolved to fiber-optic internet connections being available to more than a third of Finnish households. [1, pp. XIX-XX] [2] In manufacturing industry, laser cutting and welding have already decades ago become commonplace along the more traditional methods of production [3, pp. 335-338].

In the near future, new types of semiconductor lasers will enable improved facial recognition as laser-based components currently under development find their way into mobile devices [4]. The progress in reliability of autonomous cars and drones in turn creates a growing need for light detection and ranging (LiDAR) components [5, p. 56]. In computer science, traditional silicon-based integrated circuits (IC) are shrinking to their physical size limits. The stacking of optical data transfer capabilities to these components as well as replacement of conventional transistors using nanophotonics are under development to remedy the situation. [6, pp. 14, 22]

In addition, lasers are in key role as the research in quantum computing progresses, both in laser cooling and measurement of the quantum systems [7] [8].

One of the most rapidly growing branches of laser technology are medical devices, the global market of which is predicted to reach 2.75 billion USD in 2020 [9]. Applications in this field include, among many others, tattoo removal, brain surgery, diagnosis of disease utilizing e.g.

spectroscopy and cancer treatment by photodynamic therapy [10, pp. XXIII-XXV].

The tasks of this work is to develop methods and equipment for characterization and calibration of an ophthalmic, medical laser system. The work is conducted in a laser company (later referred to as COMPANY) for a laser system currently under development.

The aim is to both enable safe and reliable operation of this product, as well as allow for "plug and play" replacement of either of the system’s key components. Such a setup has not yet existed, and developing one is crucial for enabling effective and repeatable manufacturing of the product.

This thesis is divided into eight chapters. Laser fundamentals and electrical characteristics are explored in Chapter 2, while regulatory challenges posed by a product’s intended use for medical purposes are discussed in Chapter 3. Further properties related to laser systems and actual driving of laser diodes, as well as the need for accompanying electronics are inspected in Chapter 4.

(12)

To fully understand the task at hand, the construction of the actual laser system and necessary characterization and calibration processes are identified in Chapter 5. This includes the requirements and accuracies of said processes, in order to make sure the target parameters for the laser system are met, even if either of the system’s components is replaced.

This functionality is found to require characterization of the laser system’s measurement electronics. To validate the chosen methods for characterizing these components, a test apparatus is designed and built. Initial testing with this appliance and other laboratory equipment is performed in Chapter 6, indicating the suitability of the chosen methods and electronics solutions. Based on these promising results, an automated characterization setup is developed to replace the test apparatus. This new setup is built around an embedded computer, and runs software both written for this work and developed in COMPANY. The solution is then used for final testing, the results of which are analyzed in Chapter 7. Finally, Chapter 8 is for conclusion.

(13)

2 LASER DEVICE

By definition of an international standard, a laser device is any product or assembly incorporating a laser or laser system [11, p. 18]. In today’s world these products can range anywhere from handheld, low power laser pointers through medical laser devices to automatic missile defense systems. As can be expected, requirements and regulations differ vastly within this enormously varying field.

Key properties of a laser product are its optical output power and waveform, wavelength, and ambient operating conditions. Depending on application, the output powers of laser products may vary many orders of magnitude. For example, a typical value for a laser pointer is below 1 mW, while a medical laser device might have an output of 3 W. Moreover, current defense systems can produce beams of at least 30 kW [12], while the most powerful lasers developed have achieved powers as high as 10 PW [13, p. 1]. These extremely high powers only last for a remarkably short time, however. The highest powers currently exist only for a pulse no longer than 30 fs [14, p. 5], while the relatively low power applications of up to 10² W can usually be operated continuously.

The human eye is able to see wavelengths of electromagnetic radiation in the range of 380 to 780 nm [15, p. 2]. While this is a good target for e.g. pointers or stage lights, in many purposes it is not ideal. For example, in the case of higher power automotive range finders, it is important to use wavelengths that are not readily absorbed in the retina. For these applications near- infrared (NIR) wavelengths 905 and 1550 nm are commonly used, yet neither is indisputably considered eye safe. However, the use of this wavelength region does enable the detection of objects through some visibility impairments such as rain or fog [16, pp. 1-2].

Finally, the intended use environment sets its own requirements on the laser product. Similarly to the case of any embedded device, operating conditions are hugely different for a handheld, vehicle mounted or astronautical system, for example.

2.1 Laser fundamentals

A laser is, by definition, any kind of device able to produce or amplify electromagnetic radiation in the wavelength range from 180 nm to 1 mm primarily by utilizing controlled stimulated emission [11, p. 17]. This is also apparent from the acronym LASER – Light Amplification by Stimulated Emission of Radiation. In short, a laser works similarly to an electronic amplifier with a positive gain, except in this respect the amplified parameter is not a voltage or current, but light.

(14)

While the exact method of achieving the output radiation varies greatly among different technologies, the basic structure of a laser usually resembles some kind of an elongated cavity. The dimensions and shape of this structure can vary enormously, from lengths below 10 µm to as large as 6.5 km. [17, p. 14] A common form for a semiconductor laser is a cuboid with a volume of about 1 mm³, while a typical fiber laser comprises a thin yet flexible cylinder even many meters in length.

The production of laser light starts with the generation of the first photons in the cavity – the active medium of the laser. Optical feed-back is produced when these photons are reflected back and forth in the cavity from mirrors which it has on either end: a fully reflecting mirror at one end and a partially reflecting mirror at the other. Over time the intensity of the optical field created by the photons increases. The partially reflecting mirror subsequently lets more and more photons escape, thus forming the output of the laser, i.e. the laser beam. A steady state is eventually produced once the optical gain of the system is balanced by the loss of photons from the output mirror. [18, ch. 3.0]

2.2 Stimulated emission

Stimulated emission is the basis of positive optical gain. It is a process in which a particle or a molecule, usually an electron, in an excited state is made to relax to a lower energy state by an interaction with an incident photon. To first get to a higher energy state, the electron can for example absorb a photon, as illustrated in Figure 2.1a. [18, ch. 3.1]

Then, in order to act as a stimulant, the energy of the stimulating photon must be similar to the energy difference of the excited and relaxed state of the electron. This photon is not absorbed during the interaction. Instead the electron produces a similar, second photon of the same energy upon relaxation, as illustrated in Figure 2.1b. This, in essence, is the light amplification process. [18, ch. 3.1]

(a)An electron absorbs an incident photon and is excited

(b)An incident photon stimulates the excited electron to relax and produce a similar photon Figure 2.1.Stimulated emission

The mechanism requires, however, there to be an excess of excited electrons with regards to unexcited ones. Otherwise most of the electrons would just absorb the incident photons, which would result in a negative optical gain. In thermal equilibrium most of the electrons are in an unexcited state, which means that absorption is more probable than emission. Therefore, the

(15)

relative populations of the two energy levels of the electrons in the active medium of a laser must be reversed. In other words, to achieve stimulated emission, most of the electrons must first be excited. This process is calledpopulation inversion. [17, pp. 4-5]

However, this cannot in general be achieved with only the two energy states described so far.

Reason for this is that the photons generated by stimulated emission would just excite some other, low energy electron within the active medium and be absorbed. Instead, at least three states are required, the third allowing the newly relaxed electrons to lose a bit more energy.

This way radiation is able to escape from the laser cavity since the electrons are now in an unsuitable energy state to get excited by the photons. [17, p. 7]

In order to maintain laser emission, the electrons in the cavity must be constantly re-excited.

This can be achieved e.g. with optical pumping, a process by which the electrons are excited by photons of higher energy than the intended output laser radiation. As opposed to the photons generated within the laser cavity, these are projected into the cavity from outside of the laser. This commonly results in a four state system illustrated in Figure 2.2. [18, ch. 3.1]

2

1

3

4 Laser radiation Non-radiative

Non-radiative Pumping

Figure 2.2.Four state laser operating principle

In this arrangement, the exposure to pump photons and their subsequent absorption by the cavity electrons occurs between states 1 and 2 of the figure. Output laser radiation in turn is emitted between states 3 and 4. The transitions between states 2 and 3 as well as states 4 and 1 are non-radiative: instead of emitting photons, the electrons quickly lose energy as vibrations of the system material, called phonons. As described above, this enables the photons to escape the cavity and form a constant beam of radiation. [18, ch. 3.1]

(16)

2.3 Semiconductor laser

As opposed to solid metals, where all the valence electrons¹ of the atoms are free to move within the lattice, semiconductors contain only a certain number of free electrons. While this number is certainly greater than with insulators, it depends on temperature and of the exact material used. [18, ch. 2.1]

Gallium arsenide (GaAs) was first used as the material of laser chips instead of e.g. silicon, a group IV element. [19, p. 8] To achieve thep- andn-type semiconductor materials, silicon must be doped with group III and V elements, respectively. However, GaAs consists only of a group III metal, gallium and a group V semiconductor, arsenic. [18, ch. 2.5] Like most III/V compound semiconductors, it exhibits direct band gaps, which means that an electron can emit a photon during an energy state transition, making the material suitable for light-emitting optoelectronic components. [19, p. 7]

A simple form of a semiconductor laser is a rectangular chip with apn-junction structure that resembles a regular diode, leading to these types of devices being commonly dubbed as diode lasers. Contrary to e.g. the macroscopic chambers required for gas lasers, thepn-junction itself is able to act as the active medium of the semiconductor laser. It is commonly grown into a channel between thep- andn-layers (called claddings) of the diode in the epitaxy process.

The structure of the diode is illustrated in Figure 2.3. [18, ch. 3.3] [19, pp. 5-8]

Figure 2.3.Example semiconductor laser chip structure

In the chip, the free valence electrons on the n-side of the pn-junction reside on the conduction band, whereas the holes on thep-side represent the empty spaces for electrons on the valence band. While there can be multiple valence bands, these two are enough to effectively represent the energy transitions around the junction and to understand the principle of operation of a semiconductor laser. The junction region and these two bands are portrayed in Figure 2.4, where electrical current moves from right to left, and electrons from left to right. [1, pp. 5-6]

1Electrons on the outer, unfilled shells of an atom

(17)

Figure 2.4.Semiconductor laser pn-junction at thermal equilibrium

In the junction, the conduction band electrons transit from then-side to fill the valence band holes of thep-side. In the process they lose energy as phonons and photons, corresponding to the transitions from state 2 all the way to 1 of Figure 2.2. The pumping step from state 1 to 2 on the other hand is replaced by the electrical current resupplying the diode with new electrons.

This current is also used to achieve population inversion: when an excess of electrons is conducted to then-side of the semiconductor, the amount on the conduction band exceeds the amount on the valence band. This means that there are more electrons in a higher energy state than those in a more relaxed state, enabling stimulated emission. [18, ch. 3.3] [1, pp. 5-6]

2.4 Semiconductor laser module

While silicon electronics are usually enclosed in plastic housings with metal pins or pads, laser chips require a transparent output for the emitted light. Therefore they are enclosed in a hermetically sealed metal package with the necessary optical coupling, which can be e.g.

a fiber connector or an output lens. [1, pp. 95-96] An example of the resulting component, called a semiconductor laser module, is pictured in Figure 2.5.

Figure 2.5.Semiconductor laser module (Courtesy of COMPANY)

(18)

This procedure enables many other components to reside in the same package, such as thermoelectric cooler (TEC) modules, thermistors and photodiodes (PD) [20, p. 127]. It is even possible to enclose multiple laser chips in a single module in order to achieve higher optical output power. This is accomplished by coupling the chips’ emissions onto the same optical axis to form a single output. [21]

TECs and thermistors are particularly important sub-components because the laser diodes’

output wavelength and optical power are dependent on the temperature of the chips. As previously discussed, the non-radiative state transitions presented in Figure 2.2 transform energy of the cavity electrons into lattice vibrations, i.e. heat. Thermal expansion then alters the dimensions of the cavity, which in combination with changing band gap energy and refractive index increases the produced output wavelength. Output power in turn decreases, because charge carriers escape from the junction into the cladding material surrounding it.

There these electrons and holes are unable to contribute to the required population inversion and optical gain. [1, pp. 68, 38]

While the TEC is used to regulate temperature, the thermistor or some other kind of temperature sensor is used to provide feedback to the external circuitry and software controlling the temperature. The photodiode in turn is used to provide proportional feedback of the actual optical output of the laser by measuring the scattered radiation within the module. This way the correct operation of the laser can be ensured, even if the current draw and voltage of the module would seem appropriate. [1, p. 95] [20, p. 127]

Particularly in clinical applications, where high power laser light is aimed at biological tissues, the used treatment beams (TB) are often either invisible or not eye safe. It is necessary, however, for the operator of the device to see where they project the radiation in order to treat only the intended tissue. [10, p. 569] To solve this issue, high power laser modules often contain also a single, lower power chip emitting a different wavelength. This way, in combination with safety goggles that filter out the high power radiation, the operator can still see the low power beam imminent from the same aperture. This lower power laser is called an aiming beam (AB). [10, p. 569]

Figure 2.6 illustrates an example pinout of a multi chip module (MCM), containing a TEC, the laser diodes (LD), an aiming beam diode, a photodiode and a thermistor.

1

4 5

12

Figure 2.6.Laser module pinout

(19)

Semiconductor lasers, like many other laser types, are driven in two main modes: continuous wave (CW) and pulsed mode. While CW is often preferred for visual applications, medical treatments and military applications, pulsed operation is used e.g. for telecommunications and laser engraving. In general, the shorter the pulse, the smaller the amount of heat generated.

This allows for the laser to operate more efficiently and therefore achieve higher output powers.

[17, p. 14] [19, pp. 46, 399] [1, p. 26]

Regardless of the chosen type of output, laser diodes, like light emitting diodes (LED), are usually driven in constant current (CC) mode as opposed to constant voltage (CV) mode [22, ch. 7.4]. Also, if feedback to the laser driver is taken from the photodiode mounted within the laser module, the result is a constant power (CP) mode [20, pp. 127-129]. This kind of feedback is somewhat problematic, however. While the PD is able to provide information about the relative changes of the output power, measuring the absolute values is not possible without capturing the entire beam.

This limits the usage of CP mode to some extend, yet CV mode is by default not preferable at all. The reason behind this can be easily seen from a light radiant power, current and voltage (LIV) graph of an example laser module, pictured in Figure 2.7.

Figure 2.7.Example laser module LIV graph

As can be seen from the figure, within the optical output power range from threshold to 12 W, the drive current of the laser has a range of about 1–12 A. However, within this range the forward voltage of the laser changes only about 0.2 V. Controlling the laser output by its drive current offers therefore considerably more dynamic range than control by forward voltage.

Furthermore, since the operation principle of the diode is expressly based on the injected current, it is only sensible to use that as feedback instead of forward voltage for driving the laser.

(20)

Still, this one form of feedback is usually not enough. As discussed above, accurate temperature control of the diodes is crucial when it comes to reliable and stable operation of the laser. To demonstrate this, LIV-graphs and output wavelength spectrums were measured from an 810 nm laser module in different temperatures. These are illustrated in Figures 2.8 and 2.9.

Figure 2.8.LIV graphs of a laser module in different temperatures

Figure 2.9.Output spectrums of a laser module in different temperatures

As apparent from the figures, the drive currents required to produce the same optical output power levels are remarkably different in different temperatures. The change in output wavelength in turn is not as pronounced for this type of module used, yet it is still clearly distinguishable considering the width of the spectrum. Nonetheless, even a wavelength shift as small as 1 nm can hinder the laser’s suitability for some applications, further emphasizing the importance of accurate temperature control.

A typical semiconductor laser module containing the LDs, a PD, a TEC and a thermistor is a useful component for implementing durable and compact laser devices. While the package

(21)

contains the necessary sensors and cooling devices, these as well as the diodes themselves require suitable outside circuitry to operate. These requirements will be examined in Chapter 4.

2.5 Laser device regulation

Particularly because of the risk of permanent eye or skin damage associated with high power laser beams, laser devices are fairly heavily regulated. To aid this process, laser devices are categorized into different classes based on the hazards they pose. In Europe, these classes range from 1 to 4 [11, pp. 13-14] and in the USA from I to IV [23]. However, the international classification used in Europe is also recognized in the USA, as it is roughly equivalent with the American one.

Actual legislation then varies from country to country, referring to different standards depending on geographical location and intended use of the product. This means that, while the classification of a laser product might be identical in different parts of the world, regulation might not.

(22)

3 MEDICAL DEVICE

A medical device, according to European Union (EU) legislation, is any instrument, apparatus, software or other article that is intended by the manufacturer to be used for human beings for (certain specific) medical purposes [24, Article 2]. While for example medical gloves and pacemakers are both medical devices, their defectiveness portrays a completely different risk for the patient.

To simplify regulation, medical devices are commonly divided into three main classes, I, II and III [25] [26, Chapter I]. Moreover, to allow for more accurate risk mitigation, class II is further divided into classes IIa and IIb in the EU and Australia [24, Article 52] [27]. Alternatively, class IV is added in Brazil and Japan [28] [29]. In addition to this, the actual classification rules also differ between the regions.

3.1 Regulation

Today, medical devices are at least partially regulated in approximately 60 % of countries, and in every third a developed regulatory framework is employed. Therefore, to be allowed to market a medical device in a given country, the manufacturer is usually required to have official approval for the device in that country. [30, p. S12]

As the requirements for medical devices differ around the world, getting approvals for multiple countries has required significant resourcing from the manufacturer. To alleviate the situation, an international system of mutual acceptance has been founded by the International Electrotechnical Commission (IEC) System for Conformity Assessment Schemes for Electrotechnical Equipment and Components (IECEE). [31, p. 3]

In this system, certification bodies (CB) act as third parties between the manufacturers and the regulatory bodies of different countries and market areas, handling the necessary testing and certification. This allows for one certification process to be applicable for multiple market areas, thus decreasing the financial burden of the manufacturer. The system is called the CB-scheme. [31, p. 3]

For a medical device to be approved in this process, it needs to meet certain international standards. For example in the case of an ophthalmic¹, class 3B² medical laser device, in the perspective of laser radiation and general properties of a medical device, these standards

1Intended for diagnosis and/or treatment of eye disorders

2Here 3B is the laser device class, hence the Arabic number

(23)

would be

• IEC 60601-1:2005+AMD1:2012 – Medical electrical equipment – Part 1

• IEC 60601-2-22:2007+AMD1:2012 – Medical electrical equipment – Part 2-22

• IEC 60825-1:2014 – Safety of laser products – Part 1

• IEC 60601-2:2014 – Medical electrical equipment – Part 1-2 and

• IEC 62366-1:2015 – Medical devices – Part 1.

After this so-called CB-testing and acquisition of the required reports that validate the compliance of the device to these standards, the manufacturer can apply for approval for the device or product in a country or area.

3.2 Approval process in the USA

In the US, the approval process of a medical device depends on its class. In the simplest case of some class I or II devices, only complying to general regulatory requirements is expected. This includes e.g. general quality system regulations. However, usually a premarket notification (PMN), also called 510(k) application is required by the Food and Drug Administration (FDA). This is used to demonstrate that the medical device is both safe and effective, as well as substantially equivalent to some other approved class I/II device.

[30, pp. S14-S16]

Nonetheless, often even this is not enough. In the case that the device is of a substantially new type, or in most cases when it is a class III (highest risk) medical device, the manufacturer must apply for a premarket approval (PMA). This is a rigorous regulatory procedure intended to verify the effectiveness and safety of the medical device. [32] This process takes at least several months and it is not uncommon for it to take years.

Getting the FDA approval is the final step of the regulatory process, regardless of the application in question. After that, the manufacturer is clear to market the device in the US.

3.3 Approval process in the EU

While the FDA is a federal regulatory body responsible for handling all medical device marketing applications in USA, no such authority exists in Europe. Instead, the manufacturer is itself responsible for the application process in the EU, European Economic Area (EEA) and Switzerland. [24, Article 52]

Otherwise the application process is fairly similar. Depending on the class of the medical device, the manufacturer applies for an appropriate conformity route to fulfill the requirements in EU legislation. For certain class I devices, the manufacturer can carry out the conformity assessment by itself. However, in all other cases, a private organization certified by an EU member state is tasked at overseeing the conformity assessment process. This kind of organization is called a notified body (NB). [33] [24, Article 52]

(24)

Finally, often after months or years for a higher class device, the manufacturer affixes a European Conformity, in FrenchConformité Européenne (CE) marking on their product. This indicates its conformity with the present legislation and subsequently enables the marketing of the device. [24, Article 2]

(25)

4 LASER DEVICE CHARACTERIZATION AND CALIBRATION

In this chapter the typical construction of a laser device or system is analyzed in more detail.

The need for characterization and calibration of the sub-components and functionalities of these devices is identified. Finally, descriptions and methods of such processes are explored.

4.1 Motivation

Like any electronic devices, medical laser devices have multiple sub-components whose individual properties directly affect the output of the system. Since compliance to strict standards and specifications is required of these devices, careful characterization of said components is in order. This is a process of determining constants, such as resistance values or wavelengths, rather than evaluating some parameters across the whole operating range.

The LIV-characteristics of high-power diode lasers pose a different challenge, however. First, these properties may vary somewhat significantly from module to module, even if the chips are produced from the same wafer. Secondly, the LIV-characteristics are prone to change over time [19, p. 63]. Particularly, the produced output power of the modules decreases with respect to the supplied current during months and years of normal operation, in a process called aging.

These obstacles are illustrated in Figures 4.1 and 4.2. The first portrays the different LIV characteristics of five laser modules from the same batch, and the second how the optical output power of a laser chip slowly decreases over the course of roughly two weeks of constant operation.

The larger scale internal optical paths can differ substantially from device to device. Because of these reasons, precise laser calibration is paramount to ensure proper operation of the device. This is performed at least during production and often regularly thereafter, for example annually as preventive maintenance.

(26)

Figure 4.1.LIV-graphs of laser modules produced in the same batch (Courtesy of COMPANY)

Figure 4.2.Lifetime test of a laser chip showing aging (Courtesy of COMPANY)

In the best-case scenario, faults are detected while manufacturing the device in either the characterization or calibration process. This enables quick replacement of deficit components, saving production time and decreasing the risk of a delayed shipment.

Furthermore, particularly when it comes to complicated medical devices, failure to accomplish either of these processes may lead to inconsistent treatment results, or ultimately even patient injuries.

On the other hand, determining the operational parameters of a device is key for ensuring quality. The relevant processes must therefore be both repeatable and reliable, as well as preferably swift. While the greatest risk is a safety hazard for the user or patient, an inadequate device will always decrease the value of the manufacturer in the eyes of customers.

(27)

4.2 Laser system description

More complicated laser devices are commonly called laser systems. A typical laser system can be roughly divided into the parts illustrated in Figure 4.3. While the user interface (UI) can be as simple as an ON/OFF-switch, and the output of the system may consist of just a laser module’s fiber connector, largely more complicated variations are possible. For example, the UI can be implemented altogether on a separate device, e.g. as a mobile application. Similarly, the system’s output might contain highly sophisticated optics, such as mechanical irises and wavelength-specific mirrors. This might be the case particularly for ophthalmic instruments.

In this field, capability to adjust the output beam’s diameter while keeping output irradiance stable is often useful, if not mandatory.

Laser module Optics

Temperature controller Laser driver

Power source User interface

Cloud services Peripherals

Output monitoring

Output

Figure 4.3.Block diagram of a laser system

While the system’s UI and output stages may vary from complex to almost non-existent, some kind of laser driver circuitry is always present in a laser system. Evidently, the same applies for a power source, which is usually a battery or power supply. As for temperature controllers, the method for extracting excess heat might vary from a simple fan to intricate water cooling solutions.

The optical output power of the system is commonly monitored with a photodiode, either with one residing inside the laser module or one located closer to the actual output of the system. Finally, while not always necessary, the system is generally housed in some kind of an enclosure. This is to protect the user from the laser radiation and electric shocks, the device from dirt and debris and to enable practical transportation of the system.

These basic functionalities can be paired with numerous peripherals and different networking applications, such as foot-pedals and cloud services. Now that mobile networks have become more capable and reliable, even calibrating a customer’s device remotely is no longer at all far-fetched.

(28)

While many laser products are constructed into a single chassis, it is possible that the system consists of two or more individual parts, for example a laser source and separate output optics.

This would add a level of complexity, especially in medical devices, where requirements for reliability and safety are high. Improper integrity of the system parts might also cause the device to become expensive to repair and calibrate. This would call for the parts of the system to be interchangeable, allowing e.g. only the laser source or the output optics to be replaced.

Furthermore, passive optics are not expected to require as much calibration as laser sources, which are essentially active components. Hence, this strategy would allow for only the laser source to be sent for maintenance.

Interchangeability would also permit the manufacturer to supply its customer with a replacement part for the duration of the maintenance. This could be increasingly beneficial for example if the output optics were to be highly integrated to a larger assembly, such as a surgical robot. Dismantling such a machine only for the annual calibration of the laser would be a huge expense, which could this way be avoided.

4.3 Characterization of electronic components

As described above, in some applications parts of the laser system might be frequently swapped to another ones. Perhaps the most common of such components is the photodiode measuring the optical output power of the system, which could be replaced either by the change of the laser module or the output optics containing it. While the light-to-current response of photodiodes is highly linear, the diodes’ physical orientation inside these sub-components can significantly alter the proportional signal levels produced.

Therefore, while calibration of the whole signal loop is usually the ideal solution for a fixed laser device or system, characterization of the measurement circuit enables the attachment of any photodiode with accompanying calibration data to the system. This, again, would allow the replacement of laser system components by and on the premises of the end user, thus saving the manufacturer from a potentially expensive calibration trip.

The same principle applies for all other sensors that reside in one part of the system but are measured in another. The following sub-chapters give examples of characterization processes for a laser system incorporating a power feedback loop and output beam adjustment.

4.3.1 Photodiode current measurement

A PD is commonly used as a light sensor that linearly converts the flow of incident photons into electrical current [22, ch. 3.2]. Usually the targeted magnitude of this current is within the µA-region, so as to be both practical to measure and not disruptive with the output of the laser or system. By connecting the PD to a transimpedance amplifier, also known as a current to voltage converter (CVC), the PD acts as a negative current source between

(29)

ground and the virtual ground of the inverting input of the op-amp. As the voltage across the PD is practically 0, this configuration of photodiode operation is called photovoltaic mode.

[34, p. 409] The measurement circuit is illustrated in Figure 4.4.

Figure 4.4.PD current measurement circuit

In this circuit, the capacitor is used for stabilization and lowpass filtering. The op-amp in turn converts the current flowing through the resistor from left to right into voltage at its output by the equation

Vout =−IR. (4.1)

Since in this case the current flows from right to left, because the PD functions as a current sink, the actual resulting output voltage is positive. This can then be easily read with an analog-to-digital converter (ADC) connected to or built into a microcontroller. [35, pp. 1, 6]

There are several component values that affect how the actual current translates to the final ADC reading. These are the resistance of the resistor, the total offset voltage of the op-amp and the offset and linearity of the ADC. Since the gain of the circuit is directly dependent on the resistor, deviations in its exact value produce gain error to the resulting output voltage. For example, 1 % difference in the resistance will produce a 1 % gain error. The offsets of the op-amp and ADC on the other hand will produce offset to the measured voltage.

However, measurement of these individual parameters directly can be expected to be difficult or even impossible. Therefore, it is instead feasible to characterize the circuit as a whole. This can be done for example by the following method. First, several known negative currents are measured with the circuit. Then, theoretical readings are calculated to match these currents, based on the ideal component values. Finally, the actual and theoretical measurements are plotted together and a linear trendline is fitted to the plot.

This way, provided that the ADC has little to no non-linearities, correction factors to the errors caused by the non-ideal component values can be extracted from the trendline equation. Its slope represents the gain error and y-intercept the offset error.

For example, if the resistance would be 1 % larger than intended, the trendline’s slope should be 0.99. Then, by multiplying future results with this correction factor, they are adjusted so as to be made with an ideal resistor. Similarly, if the total offset would be 200 least significant bits (LSB) of the ADC, the y-intercept of the slope should be −200. This time, by summing

(30)

the negative value to the gain-corrected measurement, a reading matching one made with an ideal circuit is produced. This method will be explored further in Section 6.4.

4.3.2 Faster photodiode measurement

Sometimes, for example in applications where the laser output is pulsed or otherwise rapidly changing, the photodiode’s response might be too slow in the configuration described above.

To enhance the speed of this measurement circuit, a small positive bias voltage can be applied to the cathode of the PD. A practical way to achieve this is to wire the voltage to the non- inverting input of the op-amp, which then forces it to the inverting input because of the negative feedback loop. This is illustrated in Figure 4.5.

Figure 4.5.Faster PD measurement circuit

In this configuration the depletion zone in the pn-junction of the photodiode is widened, as excess electrons in the n-side cathode are drawn towards the positive bias voltage. This in turn lowers the photodiode’s capacitance and thus enables it to react faster to light level changes. It also enlarges the volume of the junction, hence increasing the probability of photon absorption. [34, p. 382]

With the applied reverse voltage, this configuration is called photocurrent operating mode.

Contrary to the photovoltaic mode described previously, where the voltage across the PD is set to zero, a so-called dark current is now formed. This is the reverse bias leakage current that remains when the light incident on the diode is reduced to zero. [22, ch. 3.3, 3.2]

4.3.3 Spot size measurement

Laser device’s output beam diameter is commonly called spot size. Adjustment of this parameter can be done e.g. by rotating a mechanical iris that directly alters the diameter of the optical path of the output beam.

To translate the iris adjustment into measurable electrical quantities, for example a potentiometer can be physically coupled to the iris. This yields a straightforward approach for measuring the displacement by first applying a known and stable voltage over the

(31)

potentiometer, then buffering the output voltage from the wiper and finally reading it with an ADC. Essentially this produces a simple voltage divider circuit, which is illustrated in Figure 4.6.

Figure 4.6. Spot size measurement circuit

To characterize the electronics in this measurement circuit, it is necessary to know precisely the reference voltage, the resistance of the series resistor R1, the total offset voltage of the op-amp and the offset of the ADC. Because of the relatively high input impedance of the op- amp, the exact resistance value of R2 is negligible. The capacitors are used for stabilization and high frequency filtering, and therefore have no practical effect to the measurement.

Similarly to the PD current measurement circuit, characterizing the individual components is not feasible. Instead, a similar approach can be used as previously. This time, instead of sinking currents from the measurement circuit, the potentiometer is disconnected, and different known voltages are applied directly to R2.

Likewise, these measurements and corresponding theoretical values are then plotted, to which a linear trendline is fitted. Correction factors to gain and offset errors are then available from the trendline’s equation. This method will be explored in practice in Section 6.4.

4.4 Calibration

As mentioned, calibration is one of the most important steps in the manufacturing process of a medical laser device. Depending on the design of such product, several different calibration phases might be required.

4.4.1 Laser diode power calibration

As discussed in Section 2.4, diode lasers and laser modules are usually driven by controlling the current flowing through them. To match software current setpoints of the device to the real power outputs, a calibration table is commonly used. Into this, several current–power

(32)

datapoints are measured along the operating range of the laser, eventually forming a discrete representation of an LI-graph. An example of such table and graphical representation of it are presented in Table 4.1 and Figure 4.7. Although the voltage of the laser diode or module is commonly monitored, it is not a particularly relevant parameter in laser systems that are driven in constant current mode. Hence it can be omitted from the calibration table.

Table 4.1. LD power calibration table

Drive Current Optical Output Drive Current Optical Output

(mA) Power (mW) (mA) Power (mW)

0 0 540 152

95 0 591 297

210 1 641 446

325 2 690 589

442 4 740 730

474 14 788 866

490 28 837 1000

505 54

Figure 4.7.Graphical representation of a LD power calibration table

While the LI-graph of a laser module is ideally perfectly linear after threshold, this is not always the case. Therefore, while a mathematical model of the curve might require less memory and fewer calculations, the table represents the actual operation of the system more accurately. However, this kind of tables can sensibly hold only a finite number of calibration points. Therefore, to produce an output with a certain power between these points, the device interpolates a current setpoint from the lower and higher value pairs in the table.

(33)

In order to ensure desired operation, the current supplied to the laser must be measured by the device. As customary for any current sensing, a small resistor is commonly wired in series with the LD and the voltage drop of this resistor is then measured with an ADC.

If these readings are paired to the actual output power values of the device, no independent characterization process for the current measurement circuitry is required. However, replacing the laser module will then require a re-calibration.

4.4.2 Main laser current, photodiode and aiming beam calibration

To produce the desired current for the laser, the microcontroller unit (MCU) handling the operation of the device commands the laser driver either over a digital bus or via an analog setpoint. Depending on driver components, a current calibration step might be required in order to prevent accidental damage to the laser. In this process, the current setpoint of the driver is calibrated to match the actual current produced by it by using a calibration table similar to the LD power calibration process. This verification is done during assembly e.g.

with an electronic load or a multimeter.

The reliability of output power monitoring of the system is at least moderately important. To achieve this, the current produced by the photodiode, or the signal of some other sensor, must be calibrated to match the actual laser output. Again, similarly to LD power calibration, a calibration table containing datapoints of output powers and PD-currents can be used.

Like the main laser source, the aiming beam requires calibration as well. This process might be just as extensive, yet often it is enough to determine a single current setting that produces a desired amount of output power from the aiming beam. This results in a simple ON/OFF-style operation.

4.4.3 Spot size calibration

In the case of an adjustable beam output, the spot size must be matched to readings taken from the adjustment appliance, e.g. a potentiometer. Again, a calibration table is a useful tool for this purpose. While the supposed voltage readings are fairly simple to acquire, this is not the case for the spot size itself.

In theory, a shade paper and a caliper could be used to measure the diameter of the output beam, yet at least for smaller diameters this is neither practical nor accurately replicable.

Instead, a beam profiler camera with a big enough sensor to accommodate the entire beam offers vastly superior capabilities for the task.

However, the output of a laser system might not be perfectly collimated, i.e. the beam might diverge or converge. For this reason, careful alignment of the beam is required. Similarly, a fixed mounting system for the camera is mandatory for the distance between the laser output and the camera to remain constant between calibrations of different devices.

(34)

5 DEVELOPMENT TASK

In this chapter, the development requirements of a characterization and calibration setup for an ophthalmic medical laser system are described. After a closer look at the system and its components, required characterization and calibration processes are recognized and necessary error limits for them chosen.

5.1 Laser system description

The system is designed to treat the fundus of the eye in combination with a photoactive drug.

It consists of two separate parts: a laser source, henceforth referred to as device, and a detachable output optics accessory called a beam shaping unit (BSU). A graphical representation of the system is presented in Figure 5.1. In this figure, power supply, UI and other peripherals of the system are not pictured for simplicity.

Figure 5.1.System description

The laser driver circuitry of the device consists of multiple parts, which are presented in Figure 5.2. The laser current is regulated by the laser diode driver (LDD), which is essentially a current source controlled by an analog input voltage. To improve the efficiency of the LDD, a pre-regulator is used between it and the power supply of the device. The pre-regulator is adjusted by an analog voltage created by a separate digital-to-analog converter (DAC). The

(35)

current and voltage of the laser module are in turn measured with a separate ADC. All of these functionalities are controlled by not just one but two microcontrollers for redundancy. A more detailed schematic of the LDD circuit is presented in Appendix A.

Figure 5.2.Simplified LDD circuitry

The output of the system is a laser beam with user definable irradiance and spot size. While the irradiance is set with a touch screen interface, the spot size is adjusted by hand via a mechanical iris in the BSU. It is located in the optical path of the light, thus cropping the beam into a desired size.

To prevent accidental illumination of a larger area of the retina than intended, the device must constantly monitor the spot setting during treatment to make sure it is not altered. This feedback is provided by a dual potentiometer connected to the iris. The device reads the wiper resistances of these two components with the circuit presented in Figure 5.3.

This results in a total of four spot size measurements, two by each ADC or channel. The doubling of the measurement allows for the device to more reliably detect any malfunctions and thus prevent patient injury. Of the components presented in the circuit, only the potentiometers are located in the BSU and everything else in the device itself.

The two parts of the laser system, the device and the BSU, are designed to be interchangeable. For this reason, the output power monitoring is implemented on the BSU in the form of a photodiode built into it. The current produced by the PD is then read by the device with a circuit presented in Figure 5.4. Again, the two separate measurements provide redundancy and increased reliability. Of the components presented in the circuit, only the photodiode is located in the BSU and everything else in the device.

(36)

Figure 5.3.Actual spot size measurement circuit

Figure 5.4.Actual photodiode current measurement circuit

Since the laser system is a medical device, the main laser is hereafter referred to as the treatment beam. This is to avoid confusion over the term laser diode, which both the treatment beam and aiming beam are produced with.

(37)

While these beams of different wavelengths are generated by two separate laser modules, they are combined inside the device into its single fiber output port. From there the laser light is directed via an optical fiber into the BSU, finally resulting in a homogeneous beam at the system’s output. This output, located at the end of the BSU, consists of a wavelength- specific mirror installed at a 45-degree angle. As the laser beam passes through the BSU, it is reflected from this mirror and directed into the eye of a patient.

In order to retain its individual parameters, the BSU has a small amount of built-in memory to store its own characterization and calibration data. When the system is started, the device reads this memory and adjusts its output parameters accordingly.

5.2 Requirements

Irradiance of the treatment beam at the output of the system is required by COMPANY to be within ±20 % of the value set by the user. However, measurement of optical power with current instruments is still not trivial. Particularly if the measured power has a dynamic range of several orders of magnitude, even an accuracy of only±2 % can be considered excellent for a power meter. Hence, the acceptable irradiance error is decreased to±15 %. This is the key goal of the whole characterization and calibration process. Moreover, irradiance of the aiming beam is required to be in the range of±35 % of 2.0 mW/cm² once enabled. Both of these requirements apply for all spot sizes.

The two spot size measurements made with each channel are first averaged by the device.

This brings the number of spot size readings down from four to two. To enable detection of potentiometer malfunction, these readings are not allowed to differ more than 150µm from each other in normal operation. The spot size readings themselves are required by COMPANY to be within±10 % of the actual values.

Finally, the optical output power measured using the photodiode in the BSU must be within

±50 % of the actual output power of the system. These requirements are summarized in Table 5.1 for clarity.

Table 5.1. System requirements set by COMPANY

System property Requirement

TB irradiance Within±15 % of the value set by user AB irradiance Within±35 % of 2.0 mW/cm² Spot size readings Within 150µm between channels and

within±10 % of actual spot size PD power reading Within±50 % of actual output power

The interchangeability of the two system parts, the device and the BSU, poses many challenges with regards to the necessary characterization and calibration processes. These have mostly been discussed in Section 4.2, however. Based on this analysis, the device part is deduced to require both treatment beam and aiming beam calibration, and

(38)

characterization of the electronics that measure photodiode current and spot size. The BSU on the other hand requires spot size and photodiode response calibration, as well as characterization of potentiometer total resistances and treatment beam’s transmission through the BSU. These steps are summarized in Table 5.2. In order to determine the acceptable error limits, these processes are analyzed individually in the following sections.

Table 5.2. Necessary characterization and calibration processes

Process System part Required equipment

AB power calibration Device Power meter

TB power calibration Device Power meter

TB transmission characterization BSU Calibrated TB laser

Spot size measurement Device Voltmeter,

circuit characterization Current meter,

Variable voltage source PD current measurement Device Variable current sink

circuit characterization

Potentiometer total BSU Resistance meter

resistance characterization

Spot size calibration BSU Beam profiler camera PD response calibration BSU Calibrated TB laser

5.2.1 Treatment beam irradiance accuracy

Irradiance is defined as the division of radiant flux (or optical power) L by the area of the receiving surfaceA, expressly

E= L

A . (5.1)

In this application, the latter is the output beam’s cross-sectional area at the calibrated focal length of 42.5 mm. As this is set to be the focal point of the beam, it is also the point of highest irradiance.

The desired irradiance is produced by the device based on the largest possible spot size. At startup, the device reads this value from the BSU connected to it and adjusts its optical output power based on the irradiance setting and spot size. Specifically, as the spot size values used in this system are not areas of the circular beam, but diameters, the device first calculates the maximum possible area from the diameter.

Because the iris in the BSU only crops the beam traversing through the BSU, optical power accuracy largely defines the irradiance accuracy of the system. This is ensured by the internal optics of the BSU, which enhance the beam’s homogeneity throughout its cross section.

(39)

The processes directly affecting optical power accuracy are power calibration and transmission characterization through the BSU. As discussed above, optical power measurement is relatively inaccurate. Therefore, an error rate of±5 % or better is targeted for power calibration. Measuring the relative change in optical power without and with the BSU on the other hand is expected to be fairly accurate, even if the absolute readings of the power meter were slightly erroneous. Thus, an accuracy of ±1 % is pursued for the transmission characterization.

Lastly, the measurement of maximum spot size during BSU spot calibration has direct, if smaller, effect on the produced irradiance. While the overall spot size measurement requirement is ±10 %, an accuracy of at least ±3 % should be achievable for the maximum spot size. This process chain is illustrated in Figure 5.5.

Figure 5.5. Processes responsible of TB irradiance

An error in spot size (i.e. spot diameter) corresponds to a spot area error of

eA = A−

(︄(1+ed)·d 2

)︄2

·π

A ^(5.2)

⇔e_A = (︄d

2 )︄2

·π−

(︄(1+e_d)·d 2

)︄2

·π

(︄d 2

)︄2

·π

(5.3)

⇔e_A =−e²_d−2ed, (5.4)

whereeAis the spot area error,dthe spot diameter anded the spot size error. Thus, a±3 % spot size error would result in a spot area error range of−6.09 to +5.91 %.

(40)

Therefore, the worst situations regarding the total irradiance error would result when either 1. (a) the optical power was 5 % larger than intended,

(b) transmission characterization resulted in further 1 % increase in the power, and (c) spot area was 6.09 % smaller, or

2. (a) the optical power was 5 % smaller than intended,

(b) transmission characterization resulted in further 1 % decrease in the power, and (c) spot area was 5.91 % larger.

These scenarios can be expressed as

(1+eE+)·E= (1+eL)·(1+eT)·L

(1−eA−)·A ^(5.5)

⇔(1+eE+)·E= (1+0.05)·(1+0.01) (1−0.0609) · L

A ^(5.6)

⇔e_E₊=0.12927≈12.9 % (5.7)

and

(1+eE−)·E= (1−e_L)·(1−e_T)·L

(1+e_A₊)·A ^(5.8)

⇔(1+eE−)·E= (1−0.05)·(1−0.01) (1+0.0591) · L

A ^(5.9)

⇔e_E−=−0.11198≈ −11.2 %, (5.10) where eE+ andeE− are the total irradiance errors, eL is the error of optical power, eT is the error of optical power caused by transmission and eA+ and eA− are the errors of spot area.

Therefore, the accuracies presented in Figure 5.5 will result in the irradiance error range of at most−11.2 to +12.9 %, which is well within the allowed±15 %.

5.2.2 Aiming beam power accuracy

Spot size is designed to be identical for both treatment beam and aiming beam. However, transmission through the BSU is found by earlier testing in COMPANY to be slightly smaller for the AB compared to the TB. As the error this causes to the optical power is expected to be

Characterization and Calibration Setup for a Medical Laser System