What If a Selfie Could Predict Your Life Expectancy?
'Doctors, myself included, still rely on the eyeball test—a split-second judgment of whether a patient looks robust or frail,' said Dr. Raymond Mak, a radiation oncologist and the director of clinical innovation for his department at the Dana-Farber Cancer Center, the faculty leader in AI implementation for the artificial intelligence in medicine (AIM) program at Mass General Brigham and an associate professor at Harvard Medical School, as well as a co-senior author of the study.
More from Flow Space
3 Small Things This Orthopedic Surgeon Wishes Women Would Do for a Longer Life
'That snap impression is subjective, yet it influences treatment decision making every day,' he told Flow Space.
So, researchers were curious just how well AI could help doctors diagnose. What they found was that in patients with cancer, looking biologically older than your chronological age, was linked with worse survival outcomes (on average, the FaceAge of cancer patients was about five years older than their chronological age).
According to the study, FaceAge not only revealed aging patterns invisible to the naked eye but also outperformed doctors in predicting short-term life expectancy for patients receiving palliative care.
'Our goal was to improve that judgment from a subjective glance to a reproducible, data-driven metric by developing an artificial intelligence algorithm called FaceAge,' Mak explained. 'Such a tool gives doctors the ability to assess patient health at low-cost and repeatedly over time with just a simple face photograph.'
An AI algorithm like FaceAge works by taking an image of a patient and then analyzing that image against a database of images of healthy individuals and those with cancer.
Mak and his team recently expanded their datasets to include millions of healthy individuals and over 20,000 cancer patients to develop an even more accurate FaceAge algorithm and to test AI performance across a larger and more diverse group of patients.
'Also, we are doing some technical work to understand how the algorithm performs over different conditions including things like, varying skin tone, impact of cosmetic surgery, use of make-up or different lighting conditions and facial expression… like whether someone is smiling or sad,' he added.
From there every image quantitatively produces a biological age estimation that is generated the same way every time, regardless of a clinician's experience level, fatigue or unconscious assumptions.
'Selfies for health!,' exclaimed Mak.
'When trained on a large and demographically varied set of face photos, the algorithm applies a consistent rule-set to every image, reducing the variability that creeps into one-to-one visual assessments,' he added. 'It does not replace the physician's judgement, but it does support that judgement with an objective reference point and flags when a patient's biological age appears discordant with their stated age.'
While the study did have limitations and biases—with further validation in larger, ethnically diverse and younger cohorts necessary before clinical adoption—FaceAge offers valuable prognostic insights independent of conventional clinical factors, with statistically significant results, even after adjusting for chronological age, sex and cancer type.
Mak added that doctors with access to FaceAge information have improved performance and reduced variability in predicting outcomes. 'By flagging people who are biologically older than their years, the technology could help us spot elevated risk for age-related conditions such as cancer and cardiovascular disease,' he said.
For midlife women—who are most commonly diagnosed with breast, lung and gastrointestinal cancers—AI and FaceAge could have life changing implications.
'The new AI models are a different breed than the AI of the early 2000s, now with an ability to learn and evolve,' Dr. Katerina Dodelzon, a radiologist specializing in breast imaging and an associate professor of radiology at Weill Cornell Medicine, told Flow Space. 'New advances in AI include subsets termed machine learning, which is an AI that can learn to make predictions or decisions, and its subset of deep learning, which uses artificial neural networks. The more data a machine learning model is exposed to, the better it performs over time.'
She says it can also help with:
Earlier and More Accurate Detection
Midlife women benefit greatly from early cancer detection, which improves survival rates for:
: AI can detect subtle changes in mammograms up to two years earlier than radiologists.
Lung Cancer: AI can flag early-stage nodules in low-dose CT scans.
: AI-assisted colonoscopy improves adenoma detection rate.
Personalized Treatment Plans
AI helps oncologists tailor therapies based on a patient's unique profile:
Genomic Data Analysis: AI can interpret massive genomic datasets to find actionable mutations. For example, in breast cancer, it helps identify candidates for hormone therapy, HER2-targeted therapy or immunotherapy, says Dodelzon.
Treatment Optimization: AI evaluates past patient responses to suggest optimal chemotherapy regimens, dosage and predict side effects.
Management
Remote Monitoring Tools: Wearables and AI apps can track vital signs, symptoms and treatment side effects. This supports real-time intervention and minimizes doctor visits.
AI Chatbots & Virtual Health Assistants: These can answer questions, schedule appointments and provide appointment and medication reminders.
Equity and Access
Many midlife women face healthcare disparities based on race, income or geography. And for women living in healthcare deserts, where access to care is limited, AI can:
Improve Access to Expertise: AI tools bring expert-level diagnostic and treatment planning to underserved or rural areas via telehealth.
Language and Literacy Support: AI-powered translation and plain-language medical explanations empower patients to understand and make informed choices.
For Mak and his team harnessing AI to save more lives is the ultimate goal.
They are currently developing new facial health recognition algorithms that can predict survival directly or other health conditions, in addition to conducting genetic analyses on a larger group of patients and opening two prospective studies.
'One is a clinical trial in cancer patients where we will compare FaceAge against conventional assessments of frailty in elderly patients,' said Mak. 'Second, we are about to open a healthy volunteer portal where people in the public can upload photos and get their own FaceAge estimate—and their photos will help us develop improved algorithms.'
And the future of AI in healthcare is set to be transformative, shifting the industry from reactive to highly proactive, personalized and precise. Dodelzon says rather than replacing doctors, AI will augment their capabilities.
This support will help catch conditions earlier, reduce diagnostic errors and streamline clinical decision-making. By leveraging vast datasets, AI will recommend treatment options tailored not only to clinical guidelines but also to a patient's unique biology and preferences.
Moreover, AI will take over many of the time-consuming administrative tasks that burden healthcare professionals, such as documentation, billing and charting, which allows for more meaningful patient interaction and personalized care.
'I think the current advances and the future development and promise of these tools is very exciting, with the potential to augment many of the routine detection and characterization tasks, and even more exciting to me, the potential to provide more prognostic in addition to diagnostic information,' Dodelzon said. 'But that is what they are—'tools' in our 'doctor's bag' that allow us to do more for our patients.'
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Medscape
3 days ago
- Medscape
Can 3 Questions Flag Depression? PHQ-3 Tested
A brief, 3-question version of the widely used Patient Health Questionnaire-9 (PHQ-9) may provide a faster, yet still accurate, way for physicians to screen for depression, new research suggests. In a study of more than 96,000 US adult participants, the PHQ-3 had 98% sensitivity and 76% specificity for predicting at least moderate or greater depressive symptoms. Additionally, it was 'highly correlated' with the full PHQ-9 and noninferior to a recently validated 4-item version for effectiveness — and it yielded a sensitivity of more than 90% across all ages, races, and ethnicities. Lead investigator Roy H. Perlis, MD, vice chair of research in the Department of Psychiatry at Mass General Brigham in Boston, said the team set out to identify the shortest version of the PHQ-9 that could still effectively screen for depression — and found that the PHQ-3 hit the 'sweet spot.' 'We let the data tell us what the constellation of symptoms are that best predict overall severity and how well they do at identifying depression. It's a way we might be able to get the maximum amount of information in the shortest amount of time,' Perlis told Medscape Medical News. However, he noted that he's not yet ready to endorse the PHQ-3 for use in clinical practice. 'I want to see further studies in clinical settings. But I think we made a good start towards showing it did well in a big, general sample of US adults,' Perlis said. The findings were published online on July 21 in JAMA Network Open . Shorter Attention Spans 'The PHQ-9 was not intended as a tool to diagnose depression. It's just a way to start the conversation; and although it's not perfect, it is useful,' Perlis noted. The original patient-reported PHQ screens for depression, anxiety, eating disorders, alcohol misuse, and somatization. Its 9-item depression scale (PHQ-9) became widely used to identify depression along with related symptoms such as fatigue and sleep disturbances. However, the length of the questionnaire has become burdensome for some patients, particularly those who prefer using phone apps over printed forms. 'Faced with a long list of survey questions, some individuals may be tempted to speed through or to not respond at all,' the investigators wrote. At the recent American Psychiatric Association (APA) annual meeting, clinicians in the audience at several sessions mentioned that their patients have been complaining that the PHQ-9 is too long and that they prefer data to be delivered in smaller 'chunks.' Perlis said that attitude was a big motivator for the study. 'We're in an era where we have shorter attention spans and people want things very quickly,' he said. 'Personally, I would always rather have the PHQ-9. The question we were trying to answer was: Can we ask fewer questions if we don't have the time or the space to ask about nine [items] and do almost as well?' Perlis said. The researchers assessed data from four waves of an online survey conducted from November 2023 to July 2024. In the first wave, they identified the optimal questionnaire items to be included. The four waves had a total of 96,234 participants (57% women; mean age, 47.3 years). Of these, 68% were White, 13% were Black, 10% were Hispanic or Latino, 5% were Asian, and 4% were classified as 'other.' In the full patient population, 26% had moderate or greater depressive symptoms, as measured by a PHQ-9 score ≥ 10. Follow-up Is Key After examining shortened versions of the PHQ-9 that ranged from including just one item up to eight items, the PHQ-3 with items 1 (interest), 2 (depressed mood), and 6 (self-esteem or failure) was deemed to be the 'optimal' version. It had a sensitivity of 0.98 (95% CI, 0.97-0.98) and a specificity of 0.76 (95% CI, 0.75-0.76) for moderate or greater depressive symptoms. Across all subgroups except participants aged 65 years or older, the sensitivity for the PHQ-3 was greater than 0.94. For that subgroup, it was 0.93. The PHQ-3 was also noninferior to previously reported sensitivity and specificity of the PHQ-Depression-4 in the whole study group and in all subgroups. The area under the receiver operating characteristic curve (AUROC) for predicting moderate or greater depressive symptoms was 0.83 for the PHQ-3. The AUROC for the PHQ-9 was 0.84. 'While a shortened scale cannot capture the full range of the PHQ-9, it may facilitate more widespread and efficient investigation of psychiatric symptoms in general population samples when participant burden and/or data collection expense must be minimized,' the investigators wrote. Overall, Perlis said the PHQ-3 could be a possible first-step screening tool, with more questions added as needed. 'What's most important is that people are screening for depression and, if someone screens positive, that they're following up with a more comprehensive evaluation and offering treatment if it's indicated,' he said. 'It's really about figuring out who needs follow-up.' Menu of Options? In an accompanying editorial, Kurt Kroenke, MD, of Regenstrief Institute, Indianapolis, Indiana, noted that the study had 'numerous strengths,' including its large and diverse study population, as well as several limitations. First, the criteria for calculating sensitivity and specificity did not include a structured psychiatric interview — the standard used in previous evaluations of the 2- and 4-item PHQs. In addition, because the researchers drew from a general population sample, future studies should focus on actual patients in both primary care and psychiatric settings, Kroenke suggested. He noted that, in addition to the various shortened PHQs, the 2-item and 4-item versions of the Patient-Reported Outcomes Measurement Information System (PROMIS) have also demonstrated benefit. 'Clinicians and researchers looking for ultrashort depression measures now have a menu from which to choose, which is good since one size may not fit all purposes,' Kroenke wrote. 'Expanded use of brief screeners to increase detection of depression has the potential to decrease the burden of the most prevalent mental disorder worldwide,' he added.
CBS News
4 days ago
- CBS News
Boston Children's nurse completes virtual Pan-Mass Challenge after missing race for stem cell donation
A Boston Children's Hospital nurse wasn't expecting to be in a hospital bed when she was meant to be riding in the Pan-Mass Challenge, but she was grateful it was for a life-changing reason: stem cell donation. Kristin McIntire is no stranger to long rides. The Boston nurse has completed 10 marathons and several 100-mile bike races. "I love exercise. I love being active," she said, jokingly adding that she commutes to work on her bike. "I don't know if you're familiar with Boston traffic, but it's certainly the most easy way to get around." But for McIntire, riding has always been about more than just fitness. It's personal and it's purposeful. "I've always needed to align that with a higher purpose, that I don't feel like I'm doing it by myself. So whether that's riding with a team, running with a team fundraising for a group, it's been a lot more meaningful for me," McIntire said. For years, McIntire served as a pediatric oncology nurse at Boston Children's Hospital, working on one of the most intense and emotional floors in the building. "There are other outcomes that aren't so great... where the family is leaving without their child," she recalled. "And that's really, really sad." She rides in the Pan-Mass Challenge with Pedal for Pediatrics, a team made up of coworkers that raises money for children and families facing cancer at Dana-Farber and Boston Children's. But last year, just as she was preparing for the big ride, Kristin received a life-changing message: she'd been matched as a stem cell donor. "It was honestly so crazy. I was like, wait, what?" McIntire said. She spent what would've been Pan-Mass Challenge weekend in a hospital bed in Florida, donating stem cells to a stranger with leukemia. "That was a pretty cool reason to stop training, I would say," McIntire said. Though she couldn't be on the course, McIntire found another way to finish the journey. In the weeks after her donation, she completed multiple 50-mile rides virtually. "This team directly provides funding that benefits those patients directly, whether that's ride share services or meal coupons, anything that can help support this family," McIntire said. McIntire has never met the person she helped. She doesn't even know their name. But she knows they're still alive. "I saved a person's life, and I was a little tired after the donation, and that was it," McIntire said. "Whether you're riding, whether you're volunteering, or whether you're out there cheering— all of that matters, and it's very important work," she added. McIntire's stem cell procedure was not a traditional bone marrow surgery. Most donations today are done through a simple outpatient blood draw called peripheral blood stem cell donation. You can learn more or join the registry here.
Yahoo
5 days ago
- Yahoo
Dana Selects Planisware for Enterprise-Wide Program Management Amid Transformative Shift in the Automotive Industry
SAN FRANCISCO, July 29, 2025--(BUSINESS WIRE)--Planisware, a global leader in strategic portfolio management software, is pleased to announce its selection by Dana Incorporated, a global leader in drivetrain and e-Propulsion systems, to implement Planisware Enterprise as their program-wide project portfolio management (PPM) tool. "Dana's decision to partner with Planisware is a testament to our continued expansion in the automotive industry, where we are increasingly recognized as a critical solution provider in the complex realm of program delivery," said Vincent Batteur, Chief Revenue Officer of Planisware. "At a time when automotive suppliers are pressed to accelerate innovation and efficiency, Planisware proves to be an invaluable ally in driving their success." The Planisware solution ensures consistency, enhances visibility, and serves as the backbone for embedding global launch processes and centralized program tracking, providing a cohesive platform to streamline communication and foster collaboration. Dana will be implementing advanced features within Planisware to manage opportunities, cost buildup, and business case analysis, enabling the making of informed, strategic decisions quickly and effectively. About Planisware Planisware is a leading business-to-business ("B2B") provider of Software-as-a-Service ("SaaS") in the rapidly growing Project Economy. Planisware's mission is to provide solutions that help organizations transform how they strategize, plan and deliver their projects, project portfolios, programs and products. With more than 700 employees across 14 offices, Planisware operates at significant scale serving circa 600 organizational clients in a wide range of verticals and functions across more than 30 countries worldwide. Planisware's clients include large international companies, medium-sized businesses and public sector entities. Planisware is listed on the regulated market of Euronext Paris (Compartment A, ISIN code FR001400PFU4, ticker symbol "PLNW"). For more information, visit and connect with Planisware on LinkedIn. View source version on Contacts Investor Relations: Benoit d'Amé +33 6 75 51 41 47 Media: Brunswick GroupHugues Boëton / Tristan Roquet Montégonplanisware@ +33 6 79 99 27 15 / +33 6 37 00 52 57 Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
