Going with our gut: backing two trailblazing healthcare startups

谈到医疗保健投资，美光风投团队才刚刚起步. 我们的团队一直在寻找在各种问题领域应用人工智能和机器学习(ML)方面真正具有创新性的初创公司. We are fascinated by the endless possibilities in healthcare & 生命科学.

Our mission is simple: Unleash technology's most innovative startups. Investments in healthcare are a newer focus for 微米, 许多企业家问我们, “Why does a semiconductor 公司 want to invest in us?” As a technology leader powering some of the biggest data transformations, we at 微米 believe in genuine innovation through AI— 和 we are discerning, but steadfast in supporting groundbreaking work with potential to change, 甚至拯救生命.

最近, 我们宣布投资两家专注于胃肠道健康创新的医疗保健初创公司. DayTwo利用微生物组数据与人工智能相结合来管理慢性代谢疾病(如糖尿病), 糖尿病前期, 和 clinical obesity) through food instead of expensive pharmaceuticals. Iterative Scopes brings cutting-edge precision medicine to gastroenterology, changing the game for endoscopic video analysis, which in its current state is known for being subjective; a costly matter when making decisions about health.

三分之一的美国人患有前驱糖尿病

You read that right: 88M adults in the US have 糖尿病前期. 34M Americans have diabetes, which constitutes 10.占美国人口的5%. 成本并不便宜:每7美元的医疗费用中就有1美元花在糖尿病及其并发症上¹. Many companies in the space are looking to tackle metabolic diseases, 但大多数方法使用混合药物，同时还要求人们坚持消除或严格限制碳水化合物摄入的饮食. This typically means radical changes to dietary habits. 然后呢?? People struggle to adhere to the guidance they’re given. 习惯很难改掉.

如果食物就是答案呢?

Imagine the possibilities if we weren’t expected to give up the foods we loved, 而是学会了如何添加食物来帮助我们达到所需的血糖平衡? DayTwo正在将人工智能应用于微生物组数据和其他生物标志物的组合，以帮助人们通过食物而不是昂贵的药物来控制糖尿病. Microbiome sampling 和 analysis is a hot topic in medicine; re搜索ers are studying the benefits of how such samples can inform health diagnosis 和 care. The microbiome is incredibly complex, comprising more than 100 trillion organisms. Analysis of a person’s gut flora could take lifetimes; but the combination of advanced computing hardware combined with AI has reduced this task to minutes.

DayTwo’s science is living proof that food is medicine. 该计划使人们能够对他们现有的饮食进行微调，以保持血糖在健康范围内. People respond differently to the same food. Let’s say you 和 your best friend meet up for breakfast every morning. 这是一种仪式，它会给你带来快乐. You eat the same dish —but your individual blood sugar levels will vary. Maybe your glucose will spike 和 your friend’s will not. Instead of advising you to cut that favorite meal out of your diet completely, DayTwo could suggest minor tweaks such as adding walnuts to yogurt, 或者黄油配米饭, that will neutralize your blood glucose spikes — all based on your unique microbiome.

美光为什么要投资?

DayTwo是唯一已知的基于微生物组的解决方案，在临床和现实世界环境中都证明了可持续的结果. 而这仅仅是个开始. The gut microbiome impacts a multitude of health conditions such as diabetes, 糖尿病前期, 肥胖和其他疾病. 进步是如此丰富，在不久的将来，当你去做例行健康检查, they just might collect a sample of your microbiome along with blood 和 urine.

To predict blood sugar responses to any food, 科学家需要一个高分辨率的微生物组视图和一个足够大的数据集来做出准确的预测. DayTwo拥有世界上最大和最丰富的专有肠道微生物组数据集，并有巨大的机会应用其科学和知识产权来改善其他疾病的治疗结果. The National Institutes of Health’s (NIH) “All of Us” initiative aims to collect data, 包括微生物组样本, from more than 1 million people to benefit precision medicine. 美国国立卫生研究院最近引用DayTwo的科学作为其精确营养研究的基础. To learn more about DayTwo’s innovative microbiome solution, visit their 网站.

Detecting the second most deadly cancer in the US is highly subjective

Colon cancer is the second most common cause of cancer deaths in the US². It is expected to increase by 60% to more than 2.200万新发病例.到2030年，癌症死亡人数将达到100万³. Endoscopies are the gold st和ard for screening for colorectal cancer, but during screening 20% of polyps are missed⁴. 目前的内窥镜手术主要基于临床直觉，产生高度主观的结果, leading to wide variations in interpretation. Endoscopies are subjective 和 there are many types of abnormal polyps, which can be quite small 和 easy to miss with the human eye. If the clinician misses something as they are doing the procedure, 没有办法回去找它了.

炎症性肠病(IBD)是指消化道慢性炎症的总称，包括克罗恩病和溃疡性结肠炎, which collectively impact approximately 3M people in the US. 病情的严重程度是用一种叫做梅奥内窥镜评分(MES)的评分技术来衡量的。. 进行内窥镜检查的医生或审查内窥镜检查录像的医生可使用此评分指数对疾病活动进行评分. It was developed in 1987 to assess initial disease severity, 活动随时间的变化, 以及对治疗的反应. 不幸的是, 这种评分技术是高度主观的，并不能真正捕捉到病人所经历的疼痛.

用直觉交换数据支持的洞察力

Outst和ing advancements have been made in oncology 和 infectious diseases, 为什么不是胃肠病学? Iterative Scopes has developed a polyp detection algorithm, SKOUT, that assists clinicians in identifying hard to see polyps, 提高决策的准确性. SKOUT在息肉周围放置一个包围框，以便在手术过程中实时引起临床医生的注意. It’s a vast oversimplification to call this technology GPS for the gut, but if you’ve ever been on the road 和 thought to yourself, ‘I never would have found this without GPS,’ there’s definitely something to the analogy.

当谈到IBD, 对严重程度的评分是如此主观，以至于错失了许多临床试验候选人的机会. 由于它是一种综合征，而不仅仅是一组症状，所以很难监测各种治疗方法的疗效, so both clinical trial selection 和 treatment analyses are extraordinarily subjective. 迭代范围可以自动解释结肠镜检查视频，可以帮助临床试验的患者选择和了解药物功效.

Iterative Scopes team featured in the image above

美光为什么要投资?

Colonoscopy videos are not usually recorded or stored; however, 迭代范围扩展了数据获取的规模，以产生此类注释数据的最大集合. 这很重要，因为机器学习技术在很大程度上依赖于可靠的训练数据来实现更高的准确性. 迭代范围专注于结肠直肠癌和IBD，并有可能影响大量人群. 胃肠病学并没有像其他学科那样利用机器学习来实现精准医疗. 迭代范围拥有团队和顾问委员会的力量，并结合了制药合作伙伴，例如 礼来公司 和 JJDC, 正因为如此, 我们相信，他们拥有综合的技术和智囊团，可以在胃肠病学领域产生有意义的影响. To learn more about Iterative Scopes, visit their 网站.

在一天结束的时候，肠道是人体最后一个真正未知的领域之一. With progress being made by companies like DayTwo 和 Iterative Scopes, 我们期待着潜在的未来进展，揭示肠脑轴的深度, one of the most powerful relationships in the body.

If you are a tech-first startup leveraging AI to innovate, 看看我们的美光风险投资页面，看看我们正在寻找什么-不要羞于与我们联系!

¹ 糖尿病的费用

² Colon cancer 2nd leading cause of cancer related death

³ Global patterns 和 trends in colorectal cancer incidence 和 mortality - PubMed (nih.gov)

⁴ Miss rate of colorectal neoplastic polyps in consecutive colonoscopies