23066

طراحي ، شبيه سازي و بهينه سازي حسگر تصويربرداري مبتني بر CMOS براي كاربردهاي تصويربرداري مغز

كارشناسي ارشد

مهندسي پزشكي - بيوالكتريك

1396

1399/09/11

دكتر احمد آيت الهي

دكتر محمد عظيم كرمي

برق

مغزيكي ازمهمترين وپيچيده ترين اعضاي بدن انسان است.مسئول عملكردهاي اساسي بدن است.پايش فعاليت مغزي به منظورتشخيص اختلالات عصبي ونورولوژيكي ، بيماري ها، وبهبودفهم ماازعملكردهاي شناختي سيستم عصبي(نوروساينس) و جراحي تومورمغزي وحذف كانون صرع مهم است. تكنيك هاي مختلفي براي پايش ساختاروعملكردمغزوجوددارد هركدام ازآنهامزايا ومعايبي دارند.دراين بين طراحي سيستم هاي تصويربرداري به منظور تصويربرداري مغزكه كم هزينه، بدون استفاده ازپرتوهاي يونيزه كننده، قابل حمل وكم حجم با توان مصرفي پايين ازاهميت ويژه اي برخورداراست كه سيستم هاي تصويربرداري نوري بسياري ازويژگي هاي گفته شده رادارند.كه ازجمله چالش هاي اصلي براي تراشه هاي تصويربرداري به كاررفته در ابزارهاي تصويربرداري نوري عصبي گستره ديناميكي ،توان مصرفي ، نسبت سيگنال به نويز، قدرت تفكيك زماني ومكاني است لذا هدف طراحي حسگرتصويربرداري باپارامترهاي ذكرشده بهبود يافته دربالا است. دراين پروژه براساس اهداف ذكرشده دوطرح پيشنهادي حسگرتصويربرداري ديجيتال مبتني برمدولاسيون فركانس پالس براي كاربرددرسيستم هاي تصويربرداري عصبي مطرح شدكه درطرح اول علاوه بردستيابي به گستره ديناميكي وسيع ،به دليل استفاده ازتكنيك هاي زيرآستانه ، self-biased وbulk-driven توان مصرفي پايين وولتاژپايين بدست آمدكه باشبيه سازي پسا لي اوت درفناوري 0.18µm CMOS ومنبع تغذيه 0.6V منجربه گستره ديناميكي 152dBوتوان مصرفي 11.25nWشد.وبراي طرح دوم كه براساس اشميت تريگراست پس ازشبيه سازي پسا لي اوت وولتاژ تغذيه 0.7V، گستره ديناميكي 144dBوتوان مصرفي63.8 nW بدست آمد.وهمچنين براي هردوطرح پيشنهاي به علت مبدل سطح پيكسل وانتقال ازحوزه ولتاژبه حوزه فركانس حساسيت كمترنسبت به نويز،به تغييرات منبع تغذيه ونسبت به پارامترهاي فرآيندمشاهده شدوازنظرپياده سازي ساده بودندهرچندمشكل اصلي درح هاي ارائه شده ضريب پرشدگي بود.درنهايت نتايج شبيه سازي طرح هاي پيشنهادشده دراين پروژه باحسگرهاي طراحي شده(كارهاي روزانجام شده) براي اين موضوع مقايسه شدكه نتايج بدست آمده وويژگي هاي طرح هاي ارائه شده پتانسيل استفاده ازآنهارابراي سيستم هاي تصويربرداري عصبي رانشان دادوپيشنهاداتي براي كارآينده ارائه شده ونتيجه گيري شد.

1399/11/01

Design,simulation and optimization of a CMOS image sensor for brain imaging applications

12/1/2020 12:00:00 AM

The brain is one of the most important and complex organs in the human body. It is responsible for the basic functions of the body. Monitoring Activity Brain in order to diagnose neurological and neurological disorders, diseases, and to improve our understanding of the cognitive functions of the system Neuroscience, as well as in applications of computer brain interface systems and brain tumor surgery and focal resection Epilepsy is important. To understand the structure and function, the brain must be Monitored in clinical situations or in normal conditions such as daily activities.This monitoring should help detect these natural conditions. Different techniques To monitor the structure and function of the brain, each of them have important advantages, disadvantages and challenges, including the resolution, high cost, bulky and other problems.In the meantime, designing imaging systems in order to Low-cost brain imaging, without ionizing radiation, portable and low volume with low power consumption, Which can be used for both semi-invasive and non-invasive imaging and therapy And can be used for applications The neural imaging system is of particular importance, that optical imaging systems have many of the features mentioned and usable in the proposed applications.Which the main challenges For imaging chips used in neural imaging tools are Dynamic range, power consumption, Signal to noise ratio, temporal and spatial resolution,But at the level of the optical imaging system itself, there are some challenges,This includes overcoming the scattering and absorption of light in the tissue, which limits the penetration depth and reduces the resolution,Therefore project goal is designing of the image sensor(imaging chip) with the above mentioned parameters improved and most importantly compatible with CMOS technology for such systems Which can be used in these neural imaging systems as implants (invasive and semi-invasive) or non-invasive in monitoring the nervous system. In this project, based on the mentioned goals, the two proposed designs of digital image sensor based on frequency modulation Pulse was introduced for use in neural imaging systems, which in the first plan, in addition to achieving Wide dynamic range, due to the use of sub-threshold, self-biased and bulk-driven techniques low power consumption and Low voltage was obtained,which by post-layout simulation in 0.18µm CMOS and 0.6V power supply,it resulted in a dynamic range of 152dB and power consumption of 11.25nW.For second plan,which is based on schmitt trriger after post-layout simulation and 0.7V supply, voltage dynamic range 144dB and power consumption was 63.8nW.Also for both proposed designs,due to the pixel level converter and transfer from the voltage domain to frequency domain less sensitive to noise,changes in the power supply and to the process parameters observed and they were simple to implement,although the main problem with proposed designs was fill factor.Finally,simulation results of proposed designs in this thesis with similar works for neural imaging was compared that obtained results and the characteristics of the proposed designs indicate the potential for their use in neural imaging systems and suggestions for the future works were presented and concluded.

پايش , حسگرتصويربرداري , نسبت سيگنال به نويز

monitoring , Signal to noise ratio