Trade shows and conferences serve numerous purposes related to business development, education, and networking, but with so many live events going virtual these days, one particular facet that we are going to focus on here is the ability to speak to industry experts face-to-face, and get to know one another beyond the ordinary exchange of phone calls and e-mails.
In our new Faces in Photonics series, we are going to profile experts whose work is reshaping business and society from all over the world, in order to learn more about them, and to let our readers get to know them a bit more as well. This first one is with Amrita Sahu, Senior Scientist at Altria.
John Lewis: Where are you from, and what got you interested in the field of imaging science?
Amrita Sahu: I grew up in east India, in the city of Kolkata. Science was an integral part of my childhood because I was surrounded by people interested in various scientific fields. My parents were my biggest influence. My dad, who is a doctor, was the best teacher in the world, and inculcated in me the love of learning. He had a special way of arousing my curiosity and used to demonstrate high school physics experiments in our kitchen!
My mother, a chemistry professor, used to often take me to her laboratory, where I met many eminent scientists. One of them was Prof. Asima Chatterjee, one of India’s most eminent scientists, who was the first woman to receive a PhD in science from an Indian University and an inventor of several antimalarial drugs. She had a deep influence on me. I still remember her telling me “The whole world is open to you, my dear.” Seeing my parents have such a positive impact on society with their work influenced me to take up science as my career.
When I came to the U.S. for my graduate studies, I was introduced to the fascinating world of imaging science. My first project was using hyperspectral and tactile imaging for identification of cancerous mammary lesions in both human and canine patients. It was a fascinating project, and we had collaborators in various hospitals and clinics testing our technology. The broader impact of such work was huge, and I completely immersed myself in it. I realized that imaging science, an interdisciplinary field involving physics, math, computer science, optics, and system development, was the way forward for me.
JL: Can you give us an analogy to help understand hyperspectral imaging?
AS: Humans and most other animals use three color receptors to see the spectrum of light—red, green, and blue. With just these three kinds of color receptors, we can see the full rainbow of this beautiful world. A creature called mantis shrimp, which are brightly colored crustaceans that live on reefs, have 12 color receptors in their eyes (see figure). With these eyes, the shrimp can distinguish between different types of coral much faster than humans could. Although this might seem like a big evolutionary win for the mantis, they are much less accurate in distinguishing different hues of the same color because human beings have a very powerful processing unit, the brain, compared to a much smaller brain of the mantis shrimp. Hyperspectral imaging combines the best of both worlds—the benefit of the optical system of mantis shrimp with multiple photoreceptors and the huge processing power of the human brain.
Hyperspectral imaging combines spectroscopy and imaging, where each image is acquired at a narrow band of the electromagnetic spectrum. Hyperspectral imaging divides the spectrum into hundreds or thousands of bands, typically covering the visible or near-infrared region. The hyperspectral imaging cube, which is three-dimensional, is simply a stack of several images. The cube contains both the spectral and spatial information of a sample. The first two dimensions are spatial (x-y axis), while the third dimension is the wavelength.
JL: What are some of the possible real-world applications of your research? Why is it important?
AS: One real-world application of my research is mammary cancer screening of human and canine patients. During my graduate studies, I was involved in the invention of a tactile imaging system that can assist doctors in the early identification of tumors in human and canine patients. This work was done in collaboration with Temple University Hospital and University of Pennsylvania Hospital doctors. Often, doctors and surgeons rely on their touch sensation to diagnose or identify the diseased regions—however, palpation is highly dependent upon the skills of the practitioner.
Malignant breast tumors are usually detected through a routine mammogram, a process that exposes the patient to potentially harmful radiation, and then confirmed by biopsy of suspected malignancies. Because those procedures require a large hospital setting with dedicated operators, a number of women have limited access to this important and often life-saving diagnostic. A patient-centric tumor screening system, which is inexpensive and simple to use, will allow many more patients to conveniently identify potential malignant tumors at early stages. Primary healthcare providers may use this device during clinical breast cancer examinations for accurately detecting small malignant tumors at early stages. This relatively simple device could be used in offices of primary care physicians where the accessibility is much greater due to proximity and convenience. This will be especially beneficial to patients in rural and remote regions, where access to a large hospital is a challenge. The fundamental discovery of the project has the potential to change the malignant tumor screening paradigm from a large hospital-centric to patient-centric model not only in the U.S., but around the globe.
I also pioneered a hyperspectral imaging method for canine mammary tumor characterization. This work was done in collaboration with the University of Pennsylvania Veterinary Hospital. Among domestic species, canines have the highest occurrence of mammary cancer. The rate of occurrence is three times that of humans. Mammary cancer in dogs has many similarities with breast cancer in women, including biology, hormone association, histological appearance, and risk factors. Mammary tumors in dogs have a wide range of biological behaviors. There are currently no available imaging methods to accurately differentiate between malignant and benign tumors pre-surgery. Therefore, surgical excision and histopathological examination of the tumors remain the standard of care in all dogs with mammary tumors. This technology can provide veterinarians with a noninvasive imaging method to determine whether surgery is necessary, or monitoring is a reasonable alternative. This is particularly useful in older dogs with concurrent health issues.
My work also has had applications in agricultural quality control. Food safety is a great public concern, and outbreaks of food-borne illnesses can lead to disturbance in society. I have invented a hyperspectral imaging-based agricultural grading system, which has been implemented by Altria (see “Hyperspectral imaging system grades agricultural products,” August 2020 issue; https://bit.ly/LFW-Hyperspectral). For this work, I have been awarded the ‘Altria Client Services John R. Nelson Innovation Award’ in 2016.
Foreign body contamination is recognized as one of the most common reasons for recall of products like food and tobacco products. Consequently, in order to comply with requirements for product safety and maintain consumer confidence, there is a need for rapid, nondestructive techniques for foreign body detection and identification in products of agricultural industry. To meet these goals, I invented a real-time contaminant detection method using hyperspectral imaging technology. Glass, metal, plastic, and foam, among others, are the most frequently cited foreign bodies in processed foods. Metal detectors are commonly implemented in food processing chains to prevent metal fragments occurring in finished products; however, these instruments are not capable of detecting other contaminants. This invention is about designing a system for real-time detection and removal of contaminants during industrial processing using hyperspectral imaging and analysis. The system incorporates a hyperspectral imaging system operating in the near-infrared region (900–1700 nm) or the shortwave-infrared region (1000–2500 nm). The hyperspectral imaging system continuously scans the sample on the conveyor belt and captures the spectral image of the sample that contains tobacco being processed and transported on the conveyor and may contain impurities like foam, cardboard, and plastic. The system works under the principle that there are subtle differences in the hyperspectral images between different materials. We have developed an intelligent algorithm to identify these differences (called ‘spectral fingerprints’) and use these to separate the impurities or undesirable materials from the sample flow in real time on the conveyor belt. Since detection of contaminants is completely automatic, it eliminates the human subjectivity and error. This technique could also be applied to the classification and identification of the common contaminants in industrial processing systems and could be used in a wide variety of industrial applications like food processing and agricultural systems, as well as other manufacturing environments.
JL: You are involved in various leadership activities outside your day-to-day work—can you briefly describe them?
AS: Yes, it has been such a privilege to have served in various leadership roles in this field. I currently serve on the Technical Committee of international conferences including IEEE and SPIE. Usually, technical committees comprise individuals with a range of expertise to span the subject area, including well-established and recognized leaders in the field from academia, government, and industry as appropriate. I have enjoyed developing technical content of conferences in the field of complex data-driven modeling and hyperspectral imaging. Being a part of such a diverse panel has taught me a great deal and has further expanded my horizons. Every year, I, as a part of these committees, ensure that well-balanced, high-quality programs are organized and presented at these conferences.
I also serve on the SPIE Early Career Award Committee. The SPIE Early Career Achievement Award is presented in recognition of significant and innovative technical contributions in the field of optical sciences. As a part of the committee, I evaluate the nominations and help decide the winner each year. It is a very rewarding experience for me to see early-career professionals being awarded for their great contributions to our field.
I have also served on the National Science Foundation panel for the last few years. This role is very special to me since my work as a panel reviewer directly contributes to basic research and people who create knowledge to transform the future. Overall, serving in such roles has enriched me with a variety of experiences that only makes me a better thinker and scientist, but also a more empathetic leader. It is incredible how many wonderful people I have met through these engagements. It is very important for me to be able to give back to the field that has given me much joy and fulfillment.
JL: You serve as an industrial advisor and mentor in several platforms—can you give us more information about these roles?
AS: I currently serve as an industrial mentor for CAMTech (Center for Arthropod Management Technologies). CAMTech is a National Science Foundation Industry/University Cooperative established more than 30 years ago to stimulate non-federal support of research and development and to accelerate technology delivery in the field of novel pest management strategies. CAMTech links the efforts of industry, government, and academia toward effective management of arthropod and nematode pests. One of the projects that I am mentoring focuses on using remote sensing, unmanned aerial vehicle (UAV) imaging, and machine learning technologies for effective pest management. It is always a great experience partnering with different academic communities and working towards our common goal of advancing this field.
Additionally, I also serve as an industrial mentor to graduate students at Virginia Tech. Mentoring graduate students is extremely rewarding, and I completely love interacting with them and helping them with their research. It reminds me of the time when I was a graduate student and was mentored by a postdoctoral scientist. Of all my roles, interacting, mentoring, and developing these bright young minds is my favorite.
JL: What makes you excited about being a scientist in this field?
AS: This is an exciting time for the evolution of this field, and I am particularly excited due to the broader social implications that machine vision technology will have on us. As hyperspectral cameras become more compact, faster, and cheaper, we will see widespread adoption of the technology in various areas such as medicine, precision agriculture, art/history preservation, environmental monitoring, biotechnology, and remote sensing, among others. When combined with real-time classification software and robotic actuators, hyperspectral imaging has the capability to revolutionize machine vision, enabling robots to perform complex tasks previously limited to humans. As the pandemic hit almost every part of world and social distancing became the new norm, we saw widespread use of imaging technologies. For example, thermal imaging cameras have been used for temperature screening at the airport and a cloud-based, AI-assisted CT imaging service has been used to detect COVID-19 pneumonia cases in China. Recently, hyperspectral imaging has been used in clinical studies to detect COVID-19 rashes. The impact of technology on society is what makes me so excited about being a scientist in this field.
JL: What are some of the challenges you faced during your career and how did you overcome them?
AS: Moving far away from my family and country to pursue my passion has been one of my biggest challenges. My parents are my biggest supporters, and they have cheered me every step of the way. I am fortunate to have a supportive spouse, who has shown tremendous respect and encouragement for my career. Over the last decade in the U.S., I have met amazing friends, coworkers, and neighbors who I consider family now. Having a good support system has helped me overcome this challenge.
Another major challenge I faced was when I transitioned from academia to industry. Initially, it was challenging to navigate and understand how the industry works, but what has helped me tremendously is to seek out knowledgeable mentors. I have sought out mentors both inside and outside my organization, and they have been instrumental in increasing not only my knowledge in the field, but also my confidence. It is natural sometimes to doubt yourself and feel insecure about your abilities, but talking it out with a mentor really helps you to see things through. As a female in a male-dominated field, I have often found myself as the only woman or the only person of color in classes, meetings, and presentations. Instead of viewing this as a challenge, I learned to embrace the difference, worked hard, and viewed it as an opportunity. Eventually, perseverance and hard work is always recognized.
JL: What advice would you give to aspiring women scientists?
AS: Aspiring women scientists should start building a professional network early. Learn to find and recognize great mentors. My mentors have inspired me and pushed me beyond what I can achieve. In this competitive, often-stressful environment, surround yourself with people who inspire you and care. A great mentor will genuinely want you to be successful.
Perseverance is key to success. It took me time to build my credibility as a scientist. Apart from focusing on technical skills, work on your leadership skills as well. It may not be obvious, but social skills are as important as technical skills to succeed. Lastly, do not be bogged down by failures. If I look back, all the challenges I faced, the failures more than the success are what gave me the strength and tenacity to ultimately be successful.
John Lewis | Editor in Chief (2018-2021)
John Lewis served as Editor in Chief of Laser Focus World from August 2018 through October 2021, after having served as the Editor in Chief of Vision Systems Design from 2016 to 2018. He has technical, industry, and journalistic qualifications, with more than 13 years of progressive content development experience working at Cognex Corporation. Prior to Cognex where his articles on machine vision were published in dozens of trade journals, he was a technical editor for Design News, covering automation, machine vision, and other engineering topics, for over six years.