AI is dynamically impacting the healthcare sector, and cancer detection is no exception. From diagnosing early-stage breast cancer to pinpointing the origin of unknown tumors, AI is solving several complex healthcare problems.

But have you ever heard of a bakery playing a role in this fight?

This article explores the incredible journey of BakeryScan, a seemingly ordinary bakery AI system developed by Hisashi Kambe’s team at BRAIN Co., Ltd. in Japan. BakeryScan’s initial purpose was far removed from the medical field, but it took an unexpected turn and became a potential game-changer in cancer cell detection.

What was the Need of Bakery AI?

Japanese bakeries needed AI for their checkout systems because of two key challenges:

1. Huge Variety of Pastries: Unlike Western bakeries with a limited selection, Japanese bakeries offer hundreds of different and constantly evolving pastries. Traditional barcode scanners wouldn’t work for this vast and ever-changing range.
2. Unwrapped Pastries: For aesthetic reasons, Japanese bakeries often display pastries unwrapped. This eliminates the option of pre-assigning barcodes, making manual identification by cashiers slow and inefficient.

AI, specifically BakeryScan, solved these problems by using cameras and deep learning to visually identify pastries based on their features (shape, color, texture) without needing barcodes. This sped up checkouts and improved customer experience.

Next Few Years of BakeryScan..

BakeryScan’s success hinges on its sophisticated AI technology. This AI can not only distinguish between similar breads but also adapt and learn over time. When unsure about an item, BakeryScan suggests possibilities to the operator, incorporating their choice to improve future accuracy. This “learning like a human” approach has propelled BakeryScan’s identification rates to a staggering 97% and beyond.

BRAIN’s innovation extends far past bread. The AI powering BakeryScan, dubbed AI-Scan, is being used in Japan to identify mixed prescription medications. Looking ahead, BRAIN aims to expand globally, potentially licensing AI-Scan to supermarkets, cafeterias, and POS system manufacturers in the US and Europe.

Bakery AI to Cancer Cell Detection

In 2017, a Doctor working for Kyoto’s Louis Pasteur Center for Medical Research saw an ad for BakeryScan. After some thinking he came to a quite funny realization that BakeryScan could possibly detect cancer cells. This sparked a collaboration between the doctor and BRAIN Co., the company behind BakeryScan.

The initial tests proved remarkably successful. BRAIN’s AI, adapted for medical applications, demonstrated a 98% accuracy rate in identifying cancerous cells. Further refinements and optimizations for analyzing large numbers of cells led to the creation of Cyto-AiSCAN.

Currently undergoing evaluation in major hospitals in Kyoto and Kobe, Cyto-AiSCAN has the potential to revolutionize cancer detection. When questioned about the underlying technology used in Cyto-AiSCAN, BRAIN Co.’s CEO, Hisashi Kambe, offered a cryptic yet intriguing response. “Original way,” he stated, followed by a smile and the enigmatic remark, “Same as bread.” This suggests that Cyto-AiSCAN might not rely solely on traditional deep learning methods, sparking curiosity about the innovative approach BRAIN Co. has developed.

Beyond the Bakery

The story of BrainCo’s AI is a shining example of how innovation often occurs in the most unexpected places. It underscores the importance of cross-disciplinary thinking and the unforeseen benefits that can arise from solving mundane problems. Today, this AI is heralding a new era in medical diagnostics, offering hope for early and more accurate cancer detection.

Checkout the role of Generative AI in Healthcare.

Summing-up

When Brain Co. made BakeryScan in 2013 to help sort pastries, they didn’t know it could become something so important. BakeryScan started small but grew into a tool that might help diagnose cancer in the future. This shows that even little ideas can lead to big changes. Science can be hard to understand, but if we keep exploring, we can figure it out.

Like BakeryScan, we can also keep trying until we reach our ultimate goal.

Let’s keep inventing for a better tomorrow.

Click here to explore more such amazing content around AI tools.