Use of topical methylene blue to image nuclear morphometry with a low-cost scanning darkfield microendoscope.

Significance: Fiber-optic microendoscopy is a promising approach to noninvasively visualize epithelial nuclear morphometry for early cancer and precancer detection. However, the broader clinical application of this approach is limited by a lack of topical contrast agents available for in vivo use. Aim: The aim of this study was to evaluate the ability to image nuclear morphometry in vivo with a novel fiber-optic microendoscope used together with topical application of methylene blue (MB), a dye with FDA approval for use in chromoendoscopy in the gastrointestinal tract. Approach: The low-cost, high-resolution microendoscope implements scanning darkfield imaging without complex optomechanical components by leveraging programmable illumination and the rolling shutter of the image sensor. We validate the integration of our system and MB staining for visualizing epithelial cell nuclei by performing ex vivo imaging on fresh animal specimens and in vivo imaging on healthy volunteers. Results: The results indicate that scanning darkfield imaging significantly reduces specular reflection and resolves epithelial nuclei with enhanced image contrast and spatial resolution compared to non-scanning widefield imaging. The image quality of darkfield images with MB staining is comparable to that of fluorescence images with proflavine staining. Conclusions: Our approach enables real-time microscopic evaluation of nuclear patterns and has the potential to be a powerful noninvasive tool for early cancer detection.

DeepDOF-SE: affordable deep-learning microscopy platform for slide-free histology.

Histopathology plays a critical role in the diagnosis and surgical management of cancer. However, access to histopathology services, especially frozen section pathology during surgery, is limited in resource-constrained settings because preparing slides from resected tissue is time-consuming, labor-intensive, and requires expensive infrastructure. Here, we report a deep-learning-enabled microscope, named DeepDOF-SE, to rapidly scan intact tissue at cellular resolution without the need for physical sectioning. Three key features jointly make DeepDOF-SE practical. First, tissue specimens are stained directly with inexpensive vital fluorescent dyes and optically sectioned with ultra-violet excitation that localizes fluorescent emission to a thin surface layer. Second, a deep-learning algorithm extends the depth-of-field, allowing rapid acquisition of in-focus images from large areas of tissue even when the tissue surface is highly irregular. Finally, a semi-supervised generative adversarial network virtually stains DeepDOF-SE fluorescence images with hematoxylin-and-eosin appearance, facilitating image interpretation by pathologists without significant additional training. We developed the DeepDOF-SE platform using a data-driven approach and validated its performance by imaging surgical resections of suspected oral tumors. Our results show that DeepDOF-SE provides histological information of diagnostic importance, offering a rapid and affordable slide-free histology platform for intraoperative tumor margin assessment and in low-resource settings.

Scanning darkfield high-resolution microendoscope for label-free microvascular imaging.

Characterization of microvascular changes during neoplastic progression has the potential to assist in discriminating precancer and early cancer from benign lesions. Here, we introduce a novel high-resolution microendoscope that leverages scanning darkfield reflectance imaging to characterize angiogenesis without exogenous contrast agents. Scanning darkfield imaging is achieved by coupling programmable illumination with a complementary metal-oxide semiconductor (CMOS) camera rolling shutter, eliminating the need for complex optomechanical components and making the system portable, low-cost (<$5,500) and simple to use. Imaging depth is extended by placing a gradient-index (GRIN) lens at the distal end of the imaging fiber to resolve subepithelial microvasculature. We validated the capability of the scanning darkfield microendoscope to visualize microvasculature at different anatomic sites in vivo by imaging the oral cavity of healthy volunteers. Images of cervical specimens resected for suspected neoplasia reveal distinct microvascular patterns in columnar and squamous epithelium with different grades of precancer, indicating the potential of scanning darkfield microendoscopy to aid in efforts to prevent cervical cancer through early diagnosis.

An integrated isothermal nucleic acid amplification test to detect HPV16 and HPV18 DNA in resource-limited settings.

High-risk human papillomavirus (HPV) DNA testing is widely acknowledged as the most sensitive cervical cancer screening method but has limited availability in resource-limited settings, where the burden of cervical cancer is highest. Recently, HPV DNA tests have been developed for use in resource-limited settings, but they remain too costly for widespread use and require instruments that are often limited to centralized laboratories. To help meet the global need for low-cost cervical cancer screening, we developed a prototype, sample-to-answer, point-of-care test for HPV16 and HPV18 DNA. Our test relies on isothermal DNA amplification and lateral flow detection, two technologies that reduce the need for complex instrumentation. We integrated all test components into a low-cost, manufacturable platform, and performance of the integrated test was evaluated with synthetic samples, provider-collected clinical samples in a high-resource setting in the United States, and self-collected clinical samples in a low-resource setting in Mozambique. We demonstrated a clinically relevant limit of detection of 1000 HPV16 or HPV18 DNA copies per test. The test requires six user steps, yields results in 45 min, and can be performed using a benchtop instrument and minicentrifuge by minimally trained personnel. The projected per-test cost is <$5, and the projected instrumentation cost is <$1000. These results show the feasibility of a sample-to-answer, point-of-care HPV DNA test. With the inclusion of other HPV types, this test has the potential to fill a critical gap for decentralized and globally accessible cervical cancer screening.

A low-cost, paper-based hybrid capture assay to detect high-risk HPV DNA for cervical cancer screening in low-resource settings.

Cervical cancer is a leading cause of cancer death for women in low-resource settings. The World Health Organization recommends that cervical cancer screening programs incorporate HPV DNA testing, but available tests are expensive, require laboratory infrastructure, and cannot be performed at the point-of-care. We developed a two-dimensional paper network (2DPN), hybrid-capture, signal amplification assay and a point-of-care sample preparation protocol to detect high-risk HPV DNA from exfoliated cervical cells within an hour. The test does not require expensive equipment and has an estimated cost of <$3 per test without the need for batching. We evaluated performance of the paper HPV DNA assay with short synthetic and genomic HPV DNA targets, HPV positive and negative cellular samples, and two sets of clinical samples. The first set of clinical samples consisted of 16 biobanked, provider-collected cervical samples from a study in El Salvador previously tested with careHPV and subsequently tested in a controlled laboratory environment. The paper HPV DNA test correctly identified eight of eight HPV-negative clinical samples and seven of eight HPV-positive clinical samples. We then performed a field evaluation of the paper HPV DNA test in a hospital laboratory in Mozambique. Cellular controls generated expected results throughout field testing with fully lyophilized sample preparation and 2DPN reagents. When evaluated with 16 residual self-collected cervicovaginal samples previously tested by the GeneXpert HPV assay ("Xpert"), the accuracy of the HPV DNA paper test in the field was reduced compared to testing in the controlled laboratory environment, with positive results obtained for all eight HPV-positive samples as well as seven of eight HPV-negative samples. Further evaluation showed reduction in performance was likely due in part to increased concentration of exfoliated cells in the self-collected clinical samples from Mozambique compared with provider-collected samples from El Salvador. Finally, a formal usability assessment was conducted with users in El Salvador and Mozambique; the assay was rated as acceptable to perform after minimal training. With additional optimization for higher cell concentrations and inclusion of an internal cellular control, the paper HPV DNA assay offers promise as a low-cost, point-of-care cervical cancer screening test in low-resource settings.

Optical imaging technologies for in vivo cancer detection in low-resource settings.

Cancer continues to affect underserved populations disproportionately. Novel optical imaging technologies, which can provide rapid, non-invasive, and accurate cancer detection at the point of care, have great potential to improve global cancer care. This article reviews the recent technical innovations and clinical translation of low-cost optical imaging technologies, highlighting the advances in both hardware and software, especially the integration of artificial intelligence, to improve in vivo cancer detection in low-resource settings. Additionally, this article provides an overview of existing challenges and future perspectives of adapting optical imaging technologies into clinical practice, which can potentially contribute to novel insights and programs that effectively improve cancer detection in low-resource settings.

Mildly dysplastic oral lesions with optically-detectable abnormalities share genetic similarities with severely dysplastic lesions.

OBJECTIVE: Optical imaging studies of oral premalignant lesions have shown that optical markers, including loss of autofluorescence and altered morphology of epithelial cell nuclei, are predictive of high-grade pathology. While these optical markers are consistently positive in lesions with moderate/severe dysplasia or cancer, they are positive only in a subset of lesions with mild dysplasia. This study compared the gene expression profiles of lesions with mild dysplasia (stratified by optical marker status) to lesions with severe dysplasia and without dysplasia. MATERIALS AND METHODS: Forty oral lesions imaged in patients undergoing oral surgery were analyzed: nine without dysplasia, nine with severe dysplasia, and 22 with mild dysplasia. Samples were submitted for high throughput gene expression analysis. RESULTS: The analysis revealed 116 genes differentially expressed among sites without dysplasia and sites with severe dysplasia; 50 were correlated with an optical marker quantifying altered nuclear morphology. Ten of 11 sites with mild dysplasia and positive optical markers (91%) had gene expression similar to sites with severe dysplasia. Nine of 11 sites with mild dysplasia and negative optical markers (82%) had similar gene expression as sites without dysplasia. CONCLUSION: This study suggests that optical imaging may help identify patients with mild dysplasia who require more intensive clinical follow-up. If validated, this would represent a significant advance in patient care for patients with oral premalignant lesions.

Allele-Specific Recombinase Polymerase Amplification to Detect Sickle Cell Disease in Low-Resource Settings.

Sickle cell disease (SCD) is a group of common, life-threatening disorders caused by a point mutation in the ß globin gene. Early diagnosis through newborn and early childhood screening, parental education, and preventive treatments are known to reduce mortality. However, the cost and complexity of conventional diagnostic methods limit the feasibility of early diagnosis for SCD in resource-limited areas worldwide. Although several point-of-care tests are commercially available, most are antibody-based tests, which cannot be used in patients who have recently received a blood transfusion. Here, we describe the development of a rapid, low-cost nucleic acid test that uses real-time fluorescence to detect the point mutation encoding hemoglobin S (HbS) in one round of isothermal recombinase polymerase amplification (RPA). When tested with a set of clinical samples from SCD patients and healthy volunteers, our assay demonstrated 100% sensitivity for both the ßA globin and ßS globin alleles and 94.7 and 97.1% specificities for the ßA globin allele and ßS globin allele, respectively (n = 91). Finally, we demonstrate proof-of-concept sample-to-answer genotyping of genomic DNA from capillary blood using an alkaline lysis procedure and direct input of diluted lysate into RPA. The workflow is performed in <30 min at a cost of <$5 USD on a commercially available benchtop fluorimeter and an open-source miniature fluorimeter. This study demonstrates the potential utility of a rapid, sample-to-answer nucleic acid test for SCD that may be implemented near the point of care and could be adapted to other disease-causing point mutations in genomic DNA.

Automated software-assisted diagnosis of esophageal squamous cell neoplasia using high-resolution microendoscopy.

BACKGROUND AND AIMS: High-resolution microendoscopy (HRME) is an optical biopsy technology that provides subcellular imaging of esophageal mucosa but requires expert interpretation of these histopathology-like images. We compared endoscopists with an automated software algorithm for detection of esophageal squamous cell neoplasia (ESCN) and evaluated the endoscopists' accuracy with and without input from the software algorithm. METHODS: Thirteen endoscopists (6 experts, 7 novices) were trained and tested on 218 post-hoc HRME images from 130 consecutive patients undergoing ESCN screening/surveillance. The automated software algorithm interpreted all images as neoplastic (high-grade dysplasia, ESCN) or non-neoplastic. All endoscopists provided their interpretation (neoplastic or non-neoplastic) and confidence level (high or low) without and with knowledge of the software overlay highlighting abnormal nuclei and software interpretation. The criterion standard was histopathology consensus diagnosis by 2 pathologists. RESULTS: The endoscopists had a higher mean sensitivity (84.3%, standard deviation [SD] 8.0% vs 76.3%, P = .004), lower specificity (75.0%, SD 5.2% vs 85.3%, P < .001) but no significant difference in accuracy (81.1%, SD 5.2% vs 79.4%, P = .26) of ESCN detection compared with the automated software algorithm. With knowledge of the software algorithm, the specificity of the endoscopists increased significantly (75.0% to 80.1%, P = .002) but not the sensitivity (84.3% to 84.8%, P = .75) or accuracy (81.1% to 83.1%, P = .13). The increase in specificity was among novices (P = .008) but not experts (P = .11). CONCLUSIONS: The software algorithm had lower sensitivity but higher specificity for ESCN detection than endoscopists. Using computer-assisted diagnosis, the endoscopists maintained high sensitivity while increasing their specificity and accuracy compared with their initial diagnosis. Automated HRME interpretation would facilitate widespread usage in resource-poor areas where this portable, low-cost technology is needed.

Cervical cancer prevention in El Salvador: A prospective evaluation of screening and triage strategies incorporating high-resolution microendoscopy to detect cervical precancer.

Cervical cancer remains a leading cause of cancer death for women in low- and middle-income countries. The goal of our study was to evaluate screening and triage strategies, including high-resolution microendoscopy (HRME), to detect cervical abnormalities concerning for precancer at the point of care. Women (n = 1824) were enrolled at the Instituto de Cáncer de El Salvador. All underwent screening by both human papillomavirus (HPV) testing using careHPV and visual inspection with acetic acid (VIA). Screen-positives, along with 10% of screen-negatives, were invited to return for a follow-up examination that included triage with VIA, colposcopy and HRME imaging. Biopsies were taken of any abnormalities identified. If no abnormalities were identified, then the worst scoring site by HRME was biopsied. The sensitivities of HPV testing and VIA to screen for cervical intraepithelial neoplasia Grade 2 or more severe diagnoses (CIN2+) were 82.1% and 75% (P = .77), while the specificities were 90.4% and 80.9% (P < .001), respectively. The sensitivities of VIA, colposcopy and HRME as triage tests for CIN2+ were 82.1%, 82.1% and 71.4%, respectively (P ≥ .38). HRME had a significantly higher specificity (66.7%) than VIA (51.9%) (P < .001) and colposcopy (53.3%) (P < .001). When evaluating different theoretical screening and triage strategies, screening with HPV testing followed by triage with HRME would result in more women receiving appropriate care (97%) compared to screening with VIA (75%) or HPV alone (90%). Our findings demonstrate that screening with HPV is superior to VIA, and that triage with HRME imaging increases the specificity of detecting CIN2+ at the point of care in a low-resource setting.

Deep learning extended depth-of-field microscope for fast and slide-free histology.

Microscopic evaluation of resected tissue plays a central role in the surgical management of cancer. Because optical microscopes have a limited depth-of-field (DOF), resected tissue is either frozen or preserved with chemical fixatives, sliced into thin sections placed on microscope slides, stained, and imaged to determine whether surgical margins are free of tumor cells-a costly and time- and labor-intensive procedure. Here, we introduce a deep-learning extended DOF (DeepDOF) microscope to quickly image large areas of freshly resected tissue to provide histologic-quality images of surgical margins without physical sectioning. The DeepDOF microscope consists of a conventional fluorescence microscope with the simple addition of an inexpensive (less than $10) phase mask inserted in the pupil plane to encode the light field and enhance the depth-invariance of the point-spread function. When used with a jointly optimized image-reconstruction algorithm, diffraction-limited optical performance to resolve subcellular features can be maintained while significantly extending the DOF (200 µm). Data from resected oral surgical specimens show that the DeepDOF microscope can consistently visualize nuclear morphology and other important diagnostic features across highly irregular resected tissue surfaces without serial refocusing. With the capability to quickly scan intact samples with subcellular detail, the DeepDOF microscope can improve tissue sampling during intraoperative tumor-margin assessment, while offering an affordable tool to provide histological information from resected tissue specimens in resource-limited settings.

Algorithm to quantify nuclear features and confidence intervals for classification of oral neoplasia from high-resolution optical images.

Purpose: In vivo optical imaging technologies like high-resolution microendoscopy (HRME) can image nuclei of the oral epithelium. In principle, automated algorithms can then calculate nuclear features to distinguish neoplastic from benign tissue. However, images frequently contain regions without visible nuclei, due to biological and technical factors, decreasing the data available to and accuracy of image analysis algorithms. Approach: We developed the nuclear density-confidence interval (ND-CI) algorithm to determine if an HRME image contains sufficient nuclei for classification, or if a better image is required. The algorithm uses a convolutional neural network to exclude image regions without visible nuclei. Then the remaining regions are used to estimate a confidence interval (CI) for the number of abnormal nuclei per mm 2 , a feature used by a previously developed algorithm (called the ND algorithm), to classify images as benign or neoplastic. The range of the CI determines whether the ND-CI algorithm can classify an image with confidence, and if so, the predicted category. The ND and ND-CI algorithm were compared by calculating their positive predictive value (PPV) and negative predictive value (NPV) on 82 oral biopsies with histopathologically confirmed diagnoses. Results: After excluding the images that could not be classified with confidence, the ND-CI algorithm had higher PPV (65% versus 59%) and NPV (78% versus 75%) than the ND algorithm. Conclusions: The ND-CI algorithm could improve the real-time classification of HRME images of the oral epithelium by informing the user if an improved image is required for diagnosis.

A PIK3CA transgenic mouse model with chemical carcinogen exposure mimics human oral tongue tumorigenesis.

Oral cancer causes significant global mortality and has a five-year survival rate of around 64%. Poor prognosis results from late-stage diagnosis, highlighting an important need to develop better approaches to detect oral premalignant lesions (OPLs) and identify which OPLs are at highest risk of progression to oral squamous cell carcinoma (OSCC). An appropriate animal model that reflects the genetic, histologic, immunologic, molecular and gross visual features of human OSCC would aid in the development and evaluation of early detection and risk assessment strategies. Here, we present an experimental PIK3CA + 4NQO transgenic mouse model of oral carcinogenesis that combines the PIK3CA oncogene mutation with oral exposure to the chemical carcinogen 4NQO, an alternate experimental transgenic mouse model with PIK3CA as well as E6 and E7 mutations, and an existing wild-type mouse model based on oral exposure to 4NQO alone. We compare changes in dorsal and ventral tongue gross visual appearance, histologic features and molecular biomarker expression over a time course of carcinogenesis. Both transgenic models exhibit cytological and architectural features of dysplasia that mimic human disease and exhibit slightly increased staining for Ki-67, a cell proliferation marker. The PIK3CA + 4NQO model additionally exhibits consistent lymphocytic infiltration, presents with prominent dorsal and ventral tongue tumours, and develops cancer quickly relative to the other models. Thus, the PIK3CA + 4NQO model recapitulates the multistep genetic model of human oral carcinogenesis and host immune response in carcinogen-induced tongue cancer, making it a useful resource for future OSCC studies.

In vivo imaging of cervical precancer using a low-cost and easy-to-use confocal microendoscope.

Cervical cancer incidence and mortality rates remain high in medically underserved areas. In this study, we present a low-cost (<$5,000), portable and user-friendly confocal microendoscope, and we report on its clinical use to image precancerous lesions in the cervix. The confocal microendoscope employs digital apertures on a digital light projector and a CMOS sensor to implement line-scanning confocal imaging. Leveraging its versatile programmability, we describe an automated aperture alignment algorithm to ensure clinical ease-of-use and to facilitate technology dissemination in low-resource settings. Imaging performance is then evaluated in ex vivo and in vivo pilot studies; results demonstrate that the confocal microendoscope can enhance visualization of nuclear morphology, contributing to significantly improved recognition of clinically important features for detection of cervical precancer.

Prospective evaluation of oral premalignant lesions using a multimodal imaging system: a pilot study.

BACKGROUND: Multimodal optical imaging, incorporating reflectance and fluorescence modalities, is a promising tool to detect oral premalignant lesions in real-time. METHODS: Images were acquired from 171 sites in 66 patient visits for clinical evaluation of oral lesions. An automated algorithm was used to classify lesions as high- or low-risk for neoplasia. Biopsies were acquired at clinically indicated sites and those classified as high-risk by imaging, at the surgeon's discretion. RESULTS: Twenty sites were biopsied based on clinical examination or imaging. Of these, 12 were indicated clinically and by imaging; 58% were moderate dysplasia or worse. Four biopsies were indicated by imaging evaluation only; 75% were moderate dysplasia or worse. Finally, four biopsies were indicated by clinical evaluation only; 75% were moderate dysplasia or worse. CONCLUSION: Multimodal imaging identified more cases of high-grade dysplasia than clinical evaluation, and can improve detection of high grade precancer in patients with oral lesions.

Low-cost, high-resolution imaging for detecting cervical precancer in medically-underserved areas of Texas.

OBJECTIVE: Cervical cancer rates in the United States have declined since the 1940's, however, cervical cancer incidence remains elevated in medically-underserved areas, especially in the Rio Grande Valley (RGV) along the Texas-Mexico border. High-resolution microendoscopy (HRME) is a low-cost, in vivo imaging technique that can identify high-grade precancerous cervical lesions (CIN2+) at the point-of-care. The goal of this study was to evaluate the performance of HRME in medically-underserved areas in Texas, comparing results to a tertiary academic medical center. METHODS: HRME was evaluated in five different outpatient clinical settings, two in Houston and three in the RGV, with medical providers of varying skill and training. Colposcopy, followed by HRME imaging, was performed on eligible women. The sensitivity and specificity of traditional colposcopy and colposcopy followed by HRME to detect CIN2+ were compared and HRME image quality was evaluated. RESULTS: 174 women (227 cervical sites) were included in the final analysis, with 12% (11% of cervical sites) diagnosed with CIN2+ on histopathology. On a per-site basis, a colposcopic impression of low-grade precancer or greater had a sensitivity of 84% and a specificity of 45% to detect CIN2+. While there was no significant difference in sensitivity (76%, p = 0.62), the specificity when using HRME was significantly higher than that of traditional colposcopy (56%, p = 0.01). There was no significant difference in HRME image quality between clinical sites (p = 0.77) or medical providers (p = 0.33). CONCLUSIONS: HRME imaging increased the specificity for detecting CIN2+ when compared to traditional colposcopy. HRME image quality remained consistent across different clinical settings.

Development of an integrated multimodal optical imaging system with real-time image analysis for the evaluation of oral premalignant lesions.

Oral premalignant lesions (OPLs), such as leukoplakia, are at risk of malignant transformation to oral cancer. Clinicians can elect to biopsy OPLs and assess them for dysplasia, a marker of increased risk. However, it is challenging to decide which OPLs need a biopsy and to select a biopsy site. We developed a multimodal optical imaging system (MMIS) that fully integrates the acquisition, display, and analysis of macroscopic white-light (WL), autofluorescence (AF), and high-resolution microendoscopy (HRME) images to noninvasively evaluate OPLs. WL and AF images identify suspicious regions with high sensitivity, which are explored at higher resolution with the HRME to improve specificity. Key features include a heat map that delineates suspicious regions according to AF images, and real-time image analysis algorithms that predict pathologic diagnosis at imaged sites. Representative examples from ongoing studies of the MMIS demonstrate its ability to identify high-grade dysplasia in OPLs that are not clinically suspicious, and to avoid unnecessary biopsies of benign OPLs that are clinically suspicious. The MMIS successfully integrates optical imaging approaches (WL, AF, and HRME) at multiple scales for the noninvasive evaluation of OPLs.

Quantitative analysis of in vivo high-resolution microendoscopic images for the detection of neoplastic colorectal polyps.

Colonoscopy is routinely performed for colorectal cancer screening but lacks the capability to accurately characterize precursor lesions and early cancers. High-resolution microendoscopy (HRME) is a low-cost imaging tool to visualize colorectal polyps with subcellular resolution. We present a computer-aided algorithm to evaluate HRME images of colorectal polyps and classify neoplastic from benign lesions. Using histopathology as the gold standard, clinically relevant features based on luminal morphology and texture are quantified to build the classification algorithm. We demonstrate that adenomatous polyps can be identified with a sensitivity and specificity of 100% and 80% using a two-feature linear discriminant model in a pilot test set. The classification algorithm presented here offers an objective framework to detect adenomatous lesions in the colon with high accuracy and can potentially improve real-time assessment of colorectal polyps.

In Vivo Multimodal Optical Imaging: Improved Detection of Oral Dysplasia in Low-Risk Oral Mucosal Lesions.

Early detection of oral cancer and oral premalignant lesions (OPL) containing dysplasia could improve oral cancer outcomes. However, general dental practitioners have difficulty distinguishing dysplastic OPLs from confounder oral mucosal lesions in low-risk populations. We evaluated the ability of two optical imaging technologies, autofluorescence imaging (AFI) and high-resolution microendoscopy (HRME), to diagnose moderate dysplasia or worse (ModDys+) in 56 oral mucosal lesions in a low-risk patient population, using histopathology as the gold standard, and in 46 clinically normal sites. AFI correctly diagnosed 91% of ModDys+ lesions, 89% of clinically normal sites, and 33% of benign lesions. Benign lesions with severe inflammation were less likely to be correctly diagnosed by AFI (13%) than those without (42%). Multimodal imaging (AFI+HRME) had higher accuracy than either modality alone; 91% of ModDys+ lesions, 93% of clinically normal sites, and 64% of benign lesions were correctly diagnosed. Photos of the 56 lesions were evaluated by 28 dentists of varied training levels, including 26 dental residents. We compared the area under the receiver operator curve (AUC) of clinical impression alone to clinical impression plus AFI and clinical impression plus multimodal imaging using k-Nearest Neighbors models. The mean AUC of the dental residents was 0.71 (range: 0.45-0.86). The addition of AFI alone to clinical impression slightly lowered the mean AUC (0.68; range: 0.40-0.82), whereas the addition of multimodal imaging to clinical impression increased the mean AUC (0.79; range: 0.61-0.90). On the basis of these findings, multimodal imaging could improve the evaluation of oral mucosal lesions in community dental settings. Cancer Prev Res; 11(8); 465-76. ©2018 AACR.