Artificial intelligence – the subject that will not quit. From shopping to travelling to Mars, AI is at the heart of many discussions and operations throughout the world. Despite the ominous predictions in films such as iRobot, Artificial intelligence is proving to be a valuable tool in many fields.
It has demonstrated a clear ability to find patterns in data that humans simply cannot find, or at least, cannot find so quickly. This is becoming more evident in healthcare as AI detects cancers, supports surgeons and determines which patients need care first.
Government Supports AI
First and foremost, AI in healthcare is now a key focus for governments across the world. It feels like many moons ago but the US budget proposal from Donald Trump’s government was the beginning for focus on AI. The research and development aspect of the proposal outlined an emphasis on AI. Since here at First Derm we are at the forefront of Dermatology AI technology, this is when things really got interesting.
Importantly, this budget proposal highlights non-defense AI investment. A move away from a focus on weapons and defense for AI is a move that paved the way for healthcare advancements.
What does AI do in Healthcare Anyway?
You may be wondering what AI does outside of robots and those concerning films about our future. We’ll dig into how AI is integral to the future of skin care and dermatology, but first let’s look at an example in hospitals.
This recent study is a great example. AI technology was used within hospitals and tested on over 75,000 patients. With AI they were able to accurately predict hospital mortality rates of patients in order to decide the best method of care. As it stands 2% of patients in US hospitals die during the admission stage. If hospitals were able to identify patients at high risk of in-hospital mortality then this would improve decision-making and outcomes for all patients involved.
This is just one example of the kind of support and range that AI technology is able to provide. In this case it is predicting hospital mortality rates, in our own studies we focus on dermatology. More on this later…
Artificial Intelligence and Skin Care
An area of focus for us is making AI a part of our future in dermatology. Detecting skin diseases is a perfect example of using AI and its pattern-detecting skills to find skin problems. Our AI matches with millions of images to give you an accurate answer on your skin concern.
Why do we need it?
Discussions on the difficulties faced by sufferers of skin conditions and the challenges dermatologists are forced to endure are nothing new. There is a global shortage of dermatologists and they are therefore time restrained and under pressure. Globally, Skin conditions are among the most common kind of ailment, just behind colds, fatigue, and headaches. In fact, around the world it’s estimated that 25% of all treatments provided to patients are for skin conditions and that up to 37% of patients seen in the clinic have at least one skin complaint.
Due to the shortage in Dermatologists, as we all know too well, General Practitioners (GPs) are asked to pick up the slack. This can often result in misdiagnoses and GPs will typically see a success rate as low as 24% in diagnosing skin conditions unlike up to 96% accuracy for Dermatologists.
How Accurate is it?
Artificial intelligence is becoming increasingly accurate in detecting skin diseases. In many cases we are now over 80% accurate in detecting a skin disease. As an example, google performed a study on 26 diseases (we can test for 43). Not unlike our very own AI technology here at First Derm. In a paper (“A Deep Learning System for Differential Diagnosis of Skin Diseases“) and accompanying blog post, they report that it achieves accuracy across 26 skin conditions when presented with images and metadata about a patient case, and they claim that it’s on par with U.S. board-certified dermatologists.
It’s great to see Google stepping up their pursuit in healthcare technology and to have achieved accuracy on 26 skin conditions is some going. Some way to go before they reach our 43 skin conditions… Google states:
“We developed a Deep Learning System (DLS) to address the most common skin conditions seen in primary care,” wrote Google software engineer Yuan Liu and Google Health technical program manager Dr. Peggy Bui. “This study highlights the potential of the DLS to augment the ability of general practitioners who did not have additional specialty training to accurately diagnose skin conditions.”
The Technical Details
If you’re interested in how this all works then here’s the details.
Liu and Bui then went on to explain that dermatologists don’t give just one diagnosis for any skin condition — instead, they generate a ranked list of possible diagnoses (a differential diagnoses) to be systematically narrowed. This is done through subsequent lab tests, imaging, procedures, and consultations. They process inputs that include one or more clinical images of the skin abnormality and up to 45 types of metadata (e.g., self-reported components of the medical history, such as age, sex, and symptoms).
According to the research team at Google, they say they evaluated the model with 17,777 de-identified cases from 17 primary care clinics across two states. During training, the model leveraged over 50,000 differential diagnoses provided by over 40 dermatologists.
Test Our AI Technology For Free Here.
In a test of the system’s diagnostic accuracy, the researchers compiled diagnoses from three U.S. board-certified dermatologists. Just over 3,750 cases were aggregated to derive the ground truth labels, and the AI system’s ranked list of skin conditions achieved 71% and 93% top-1 and top-3 accuracies, respectively.
Furthermore, when the system was compared against three categories of clinicians (dermatologists, primary care physicians, and nurse practitioners) on a subset of the validation data set, the team reports that its top three predictions demonstrated a top-3 diagnostic accuracy of 90%, or comparable to dermatologists (75%) and “substantially higher” than primary care physicians (60%) and nurse practitioners (55%).
Liu and Bui go on to conclude their study with a fair assessment of our growing market.
“The success of deep learning to inform the differential diagnosis of skin disease is highly encouraging of such a tool’s potential to assist clinicians. For example, such a DLS could help triage cases to guide prioritization for clinical care or help non-dermatologists initiate dermatologic care more accurately and could potentially improve access [to care].”
So AI is more accurate than a Doctor?
It may seem unbelievable but compared to your average GP, the AI is generally more accurate. It’s worth remembering that it is finding patterns in thousands of images within seconds, whereas a GP has a lot of different health problems to remember!
The challenges a Doctor faces when diagnosing skin diseases has long been discussed as demonstrated by a 1981 research piece by Dr. David L. Ramsay and Dr. Alissa Benimoff Fox named ‘The ability of primary care physicians to recognize the common dermatoses’ – this research paper highlighted that Doctors achieved only a 16% accuracy in some cases whereas a Dermatologist achieved 93% (in some cases 100% accuracy).
Since 1981, Doctors have of course improved in accuracy whilst technology has advanced alongside. We’re now at a stage where it is no longer Dermatologists that we compare Doctors to, but AI.
What Diseases are easier to diagnose?
In most cases forms of Skin cancer are the main focus. In particular, Melanoma and Cancerous moles. Frequently, as demonstrated in this study, the AI intelligence is more accurate than a doctor and in fact, more accurate than a Dermatologist when testing for Melanoma.
Try our FREE Skin Image Search™
UC Santa Barbara undergrad Abhisheck Bhattacharya, who developed the melanoma project along with UC San Francisco professor, Dr. Dexter Hadley, has so far proven a 96% accuracy rate using this model.
“We’re applying computer vision to solving medical problems,” said Bhattacharya.
“What we’re trying to do is something humans can’t do,” says Dr. Hadley. “Humans can’t predict what goes on under the skin from looking from the top.”
Regardless of the accuracy rate in the laboratory, researchers at UC are taking steps toward proving the success rate in a clinical trial before the technology is introduced to the public.
At Stanford, researchers found that their app is also capable of identifying skin cancer as well as dermatologists (91% accuracy). The Stanford team says the aim of developing their program is not to replace human dermatologists. However it is an inexpensive option for early screening. Both teams say that the model neural network looks promising, however, more rigorous assessments of its safety would need to be made before such a program could go public.
Have a questions about your skin? Consult a dermatologist now.
What else does AI Dermatology work on?
After a number of in depth tests and analysis with our Dermatologists we have reached a new breakthrough with over 80% accuracy for visual STI’s. You can test it out here.
Spots, marks, warts, moles and more are easily detected by our AI technology and will provide an answer within seconds. No need to wait for an STD testing kit, simply send in the photos and get your results.
STDs are commonplace and should always be treated as swiftly as possible. Due to this demand there are many companies accommodating for sexually transmitted diseases. Various options exist online such as free testing kits, swabs and STD test boxes. However, sometimes we need the results sooner, and this is why our Free Image tester was created: to save you valuable time and put your mind at ease.
With artificial intelligence (AI), computer algorithms such as ours are learning more about acne and the specific type you have. This can all be done by a simple image upload. It will typically take under a second to know what form of acne you have that can help guide you to the right treatment that you can use when you see your dermatologist in person.
Is AI being used now?
Yes! AI is being used right now in many studies and clinics for skin concerns. Here’s a perfect example:
STOCKHOLM, SWEDEN, November 27, 2018: Medicoo, one of Sweden’s first digital healthcare centers, have launched a fully automated solution, with the help of image recognition and artificial intelligence (AI). They can now quickly screen various skin disorders, including skin cancer, malignant melanoma.
Malignant melanoma is a cancer on the rise in Sweden, and is curable if caught early enough and removed with surgery. The survival rate is 80-85 percent .
In an eye-catching feature of the TV4 News channel dated 6th November, the Swedish patient melanoma association demanded that skin cancer care to be managed nationally. They encouraged the use of modern technology such as smartphone cameras.
“Swedes have an average of 1.6 skin diseases annually, so the need for care in this area is enormous,” says Jan Lundblad, CEO of Medicoo. “There are long waiting times to a dermatologist. We want to help to streamline care with the help of technology. “
Check out your skin concern using artificial intelligence
With an ordinary mobile phone camera and artificial intelligence, skin screening can be scaled. Skin screening via Medicoo, is in collaboration with First Derm and it will initially be offered free of charge.
“Cases of malignant melanoma could be detected in time when the access to care is simple,” says Camilla Arneving, Marketing Manager.
“The AI has been trained with clinical thinking and, like a regular doctor, it becomes better over time. Research is still being carried out, but it has already achieved an acceptable level of accuracy. “Says Ann Zyto, Chief Executive Officer at Medicoo. “We are very excited to offer a free skin screening that can help save lives!”
Can I try AI on My Skin?
100% you can. You can try our AI here for free and anonymously. Here’s some do’s and don’t for our skin image searcher:
- Do use a smartphone with a good camera
- The picture size pixels Mb does not matter
- Do use good natural lighting
- Don’t have irrelevant distractions in the picture, like a watch, jewelry, pointing your finger or draw on the picture
- Do clearly capture the area of concern
- Do take focused pictures