Deep learning-aided decision support for diagnosis of skin disease across skin tones.

Although advances in deep learning systems for image-based medical diagnosis demonstrate their potential to augment clinical decision-making, the effectiveness of physician-machine partnerships remains an open question, in part because physicians and algorithms are both susceptible to systematic errors, especially for diagnosis of underrepresented populations. Here we present results from a large-scale digital experiment involving board-certified dermatologists (n = 389) and primary-care physicians (n = 459) from 39 countries to evaluate the accuracy of diagnoses submitted by physicians in a store-and-forward teledermatology simulation. In this experiment, physicians were presented with 364 images spanning 46 skin diseases and asked to submit up to four differential diagnoses. Specialists and generalists achieved diagnostic accuracies of 38% and 19%, respectively, but both specialists and generalists were four percentage points less accurate for the diagnosis of images of dark skin as compared to light skin. Fair deep learning system decision support improved the diagnostic accuracy of both specialists and generalists by more than 33%, but exacerbated the gap in the diagnostic accuracy of generalists across skin tones. These results demonstrate that well-designed physician-machine partnerships can enhance the diagnostic accuracy of physicians, illustrating that success in improving overall diagnostic accuracy does not necessarily address bias.

BioAutoMATED: An end-to-end automated machine learning tool for explanation and design of biological sequences.

The design choices underlying machine-learning (ML) models present important barriers to entry for many biologists who aim to incorporate ML in their research. Automated machine-learning (AutoML) algorithms can address many challenges that come with applying ML to the life sciences. However, these algorithms are rarely used in systems and synthetic biology studies because they typically do not explicitly handle biological sequences (e.g., nucleotide, amino acid, or glycan sequences) and cannot be easily compared with other AutoML algorithms. Here, we present BioAutoMATED, an AutoML platform for biological sequence analysis that integrates multiple AutoML methods into a unified framework. Users are automatically provided with relevant techniques for analyzing, interpreting, and designing biological sequences. BioAutoMATED predicts gene regulation, peptide-drug interactions, and glycan annotation, and designs optimized synthetic biology components, revealing salient sequence characteristics. By automating sequence modeling, BioAutoMATED allows life scientists to incorporate ML more readily into their work.

Integrated multimodal artificial intelligence framework for healthcare applications.

Artificial intelligence (AI) systems hold great promise to improve healthcare over the next decades. Specifically, AI systems leveraging multiple data sources and input modalities are poised to become a viable method to deliver more accurate results and deployable pipelines across a wide range of applications. In this work, we propose and evaluate a unified Holistic AI in Medicine (HAIM) framework to facilitate the generation and testing of AI systems that leverage multimodal inputs. Our approach uses generalizable data pre-processing and machine learning modeling stages that can be readily adapted for research and deployment in healthcare environments. We evaluate our HAIM framework by training and characterizing 14,324 independent models based on HAIM-MIMIC-MM, a multimodal clinical database (N = 34,537 samples) containing 7279 unique hospitalizations and 6485 patients, spanning all possible input combinations of 4 data modalities (i.e., tabular, time-series, text, and images), 11 unique data sources and 12 predictive tasks. We show that this framework can consistently and robustly produce models that outperform similar single-source approaches across various healthcare demonstrations (by 6-33%), including 10 distinct chest pathology diagnoses, along with length-of-stay and 48 h mortality predictions. We also quantify the contribution of each modality and data source using Shapley values, which demonstrates the heterogeneity in data modality importance and the necessity of multimodal inputs across different healthcare-relevant tasks. The generalizable properties and flexibility of our Holistic AI in Medicine (HAIM) framework could offer a promising pathway for future multimodal predictive systems in clinical and operational healthcare settings.

Wearable materials with embedded synthetic biology sensors for biomolecule detection.

Integrating synthetic biology into wearables could expand opportunities for noninvasive monitoring of physiological status, disease states and exposure to pathogens or toxins. However, the operation of synthetic circuits generally requires the presence of living, engineered bacteria, which has limited their application in wearables. Here we report lightweight, flexible substrates and textiles functionalized with freeze-dried, cell-free synthetic circuits, including CRISPR-based tools, that detect metabolites, chemicals and pathogen nucleic acid signatures. The wearable devices are activated upon rehydration from aqueous exposure events and report the presence of specific molecular targets by colorimetric changes or via an optical fiber network that detects fluorescent and luminescent outputs. The detection limits for nucleic acids rival current laboratory methods such as quantitative PCR. We demonstrate the development of a face mask with a lyophilized CRISPR sensor for wearable, noninvasive detection of SARS-CoV-2 at room temperature within 90 min, requiring no user intervention other than the press of a button.

Using deep learning for dermatologist-level detection of suspicious pigmented skin lesions from wide-field images.

A reported 96,480 people were diagnosed with melanoma in the United States in 2019, leading to 7230 reported deaths. Early-stage identification of suspicious pigmented lesions (SPLs) in primary care settings can lead to improved melanoma prognosis and a possible 20-fold reduction in treatment cost. Despite this clinical and economic value, efficient tools for SPL detection are mostly absent. To bridge this gap, we developed an SPL analysis system for wide-field images using deep convolutional neural networks (DCNNs) and applied it to a 38,283 dermatological dataset collected from 133 patients and publicly available images. These images were obtained from a variety of consumer-grade cameras (15,244 nondermoscopy) and classified by three board-certified dermatologists. Our system achieved more than 90.3% sensitivity (95% confidence interval, 90 to 90.6) and 89.9% specificity (89.6 to 90.2%) in distinguishing SPLs from nonsuspicious lesions, skin, and complex backgrounds, avoiding the need for cumbersome individual lesion imaging. We also present a new method to extract intrapatient lesion saliency (ugly duckling criteria) on the basis of DCNN features from detected lesions. This saliency ranking was validated against three board-certified dermatologists using a set of 135 individual wide-field images from 68 dermatological patients not included in the DCNN training set, exhibiting 82.96% (67.88 to 88.26%) agreement with at least one of the top three lesions in the dermatological consensus ranking. This method could allow for rapid and accurate assessments of pigmented lesion suspiciousness within a primary care visit and could enable improved patient triaging, utilization of resources, and earlier treatment of melanoma.

Computer-aided classification of suspicious pigmented lesions using wide-field images.

BACKGROUND AND OBJECTIVE: Early identification of melanoma is conducted through whole-body visual examinations to detect suspicious pigmented lesions, a situation that fluctuates in accuracy depending on the experience and time of the examiner. Computer-aided diagnosis tools for skin lesions are typically trained using pre-selected single-lesion images, taken under controlled conditions, which limits their use in wide-field scenes. Here, we propose a computer-aided classifier system with such input conditions to aid in the rapid identification of suspicious pigmented lesions at the primary care level. METHODS: 133 patients with a multitude of skin lesions were recruited for this study. All lesions were examined by a board-certified dermatologist and classified into "suspicious" and "non-suspicious". A new clinical database was acquired and created by taking Wide-Field images of all major body parts with a consumer-grade camera under natural illumination condition and with a consistent source of image variability. 3-8 images were acquired per patient on different sites of the body, and a total of 1759 pigmented lesions were extracted. A machine learning classifier was optimized and build into a computer aided classification system to binary classify each lesion using a suspiciousness score. RESULTS: In a testing set, our computer-aided classification system achieved a sensitivity of 100% for suspicious pigmented lesions that were later confirmed by dermoscopy examination ("SPL_A") and 83.2% for suspicious pigmented lesions that were not confirmed after examination ("SPL_B"). Sensitivity for non-suspicious lesions was 72.1%, and accuracy was 75.9%. With these results we defined a suspiciousness score that is aligned with common macro-screening (naked eye) practices. CONCLUSIONS: This work demonstrates that wide-field photography combined with computer-aided classification systems can distinguish suspicious from non-suspicious pigmented lesions, and might be effective to assess the severity of a suspicious pigmented lesions. We believe this approach could be useful to support skin screenings at a population-level.

A deep learning approach to programmable RNA switches.

Engineered RNA elements are programmable tools capable of detecting small molecules, proteins, and nucleic acids. Predicting the behavior of these synthetic biology components remains a challenge, a situation that could be addressed through enhanced pattern recognition from deep learning. Here, we investigate Deep Neural Networks (DNN) to predict toehold switch function as a canonical riboswitch model in synthetic biology. To facilitate DNN training, we synthesize and characterize in vivo a dataset of 91,534 toehold switches spanning 23 viral genomes and 906 human transcription factors. DNNs trained on nucleotide sequences outperform (R2 = 0.43-0.70) previous state-of-the-art thermodynamic and kinetic models (R2 = 0.04-0.15) and allow for human-understandable attention-visualizations (VIS4Map) to identify success and failure modes. This work shows that deep learning approaches can be used for functionality predictions and insight generation in RNA synthetic biology.

Creating CRISPR-responsive smart materials for diagnostics and programmable cargo release.

Materials that sense and respond to biological signals in their environment have a broad range of potential applications in drug delivery, medical devices and diagnostics. Nucleic acids are important biological cues that encode information about organismal identity and clinically relevant phenotypes such as drug resistance. We recently developed a strategy to design nucleic acid-responsive materials using the CRISPR-associated nuclease Cas12a as a user-programmable sensor and material actuator. This approach improves on the sensitivity of current DNA-responsive materials while enabling their rapid repurposing toward new sequence targets. Here, we provide a comprehensive resource for the design, synthesis and actuation of CRISPR-responsive hydrogels. First, we provide guidelines for the synthesis of Cas12a guide RNAs (gRNAs) for in vitro applications. We then outline methods for the synthesis of both polyethylene glycol-DNA (PEG-DNA) and polyacrylamide-DNA (PA-DNA) hydrogels, as well as their controlled degradation using Cas12a for the release of cargos, including small molecules, enzymes, nanoparticles and living cells within hours. Finally, we detail the design and assembly of microfluidic paper-based devices that use Cas12a-sensitive hydrogels to convert DNA inputs into a variety of visual and electronic readouts for use in diagnostics. Following the initial validation of the gRNA and Cas12a components (1 d), the synthesis and testing of either PEG-DNA or PA-DNA hydrogels require 3-4 d of laboratory time. Optional extensions, including the release of primary human cells or the design of the paper-based diagnostic, require an additional 2-3 d each.

Programmable CRISPR-responsive smart materials.

Stimuli-responsive materials activated by biological signals play an increasingly important role in biotechnology applications. We exploit the programmability of CRISPR-associated nucleases to actuate hydrogels containing DNA as a structural element or as an anchor for pendant groups. After activation by guide RNA-defined inputs, Cas12a cleaves DNA in the gels, thereby converting biological information into changes in material properties. We report four applications: (i) branched poly(ethylene glycol) hydrogels releasing DNA-anchored compounds, (ii) degradable polyacrylamide-DNA hydrogels encapsulating nanoparticles and live cells, (iii) conductive carbon-black-DNA hydrogels acting as degradable electrical fuses, and (iv) a polyacrylamide-DNA hydrogel operating as a fluidic valve with an electrical readout for remote signaling. These materials allow for a range of in vitro applications in tissue engineering, bioelectronics, and diagnostics.

Interconnected Microphysiological Systems for Quantitative Biology and Pharmacology Studies.

Microphysiological systems (MPSs) are in vitro models that capture facets of in vivo organ function through use of specialized culture microenvironments, including 3D matrices and microperfusion. Here, we report an approach to co-culture multiple different MPSs linked together physiologically on re-useable, open-system microfluidic platforms that are compatible with the quantitative study of a range of compounds, including lipophilic drugs. We describe three different platform designs - "4-way", "7-way", and "10-way" - each accommodating a mixing chamber and up to 4, 7, or 10 MPSs. Platforms accommodate multiple different MPS flow configurations, each with internal re-circulation to enhance molecular exchange, and feature on-board pneumatically-driven pumps with independently programmable flow rates to provide precise control over both intra- and inter-MPS flow partitioning and drug distribution. We first developed a 4-MPS system, showing accurate prediction of secreted liver protein distribution and 2-week maintenance of phenotypic markers. We then developed 7-MPS and 10-MPS platforms, demonstrating reliable, robust operation and maintenance of MPS phenotypic function for 3 weeks (7-way) and 4 weeks (10-way) of continuous interaction, as well as PK analysis of diclofenac metabolism. This study illustrates several generalizable design and operational principles for implementing multi-MPS "physiome-on-a-chip" approaches in drug discovery.

Analysis of white blood cell dynamics in nailfold capillaries.

Based on video data acquired with low-cost, portable microscopy equipment, we introduce a semi-automatic method to count visual gaps in the blood flow as a proxy for white blood cells (WBC) passing through nailfold capillaries. Following minimal user interaction and a pre-processing stage, our method consists in the spatio-temporal segmentation and analysis of capillary profiles. Besides the mere count information, it also estimates the speed associated with every WBC event. The accuracy of our algorithm is validated through the analysis of two capillaries acquired from one healthy subject. Results are compared with manual counts from four human raters and confronted with related physiological data reported in literature.

Efficacy and safety evaluation of systemic extremely low frequency magnetic fields used in the healing of diabetic foot ulcers--phase II data.

BACKGROUND AND AIMS: Cellular and animal models investigating extremely low frequency magnetic fields (ELF-MF) have reported promotion of leukocyte-endothelial interactions, angiogenesis, myofibroblast and keratinocyte proliferation, improvement of peripheral neuropathy and diabetic wound healing. In humans, it has also been reported that systemic exposure to ELF-MF stimulates peripheral blood mononuclear cells, promoting angiogenesis and healing of chronic leg ulcers. The aim of the study was to investigate the effect of exposing different blood volumes to specific ELF-MFs (120 Hz sinusoidal waves of 0.4-0.9 mT RMS) to induce healing of diabetic foot ulcers (DFUs). METHODS: Twenty six diabetic patients with non-responsive DFUs were divided into two exposure groups to receive treatment and record healing time. The forearm group, exposed to ELF-MF 2 h/day, twice weekly (3.6 l of blood/session); and the thorax group, exposed 25 min/day, 2 times/week (162.5 l of blood/session). Treatment period was 100 days or upon complete healing. Ulcer recurrences and adverse effects were investigated during short-term (<1 year) and long-term (3.4-7.8 years) follow-up. RESULTS: Mean healing time was 61.48 ± 33.08 days in the forearm group and 62.56 ± 29.33 days for the thorax group. No adverse effects or ulcer recurrences in the original ulcer site were reported during treatment, the short-term follow-up period or the long-term follow-up period in both groups. CONCLUSIONS: Healing time was independent of the amount of blood exposed to ELF-MF used in this trial. ELF-MFs are effective and safe and could be applied to non-healing DFUs in conjunction with other preventive interventions to reduce DFUs complications.