Ahmed Hosny

Harvard University | Harvard

Artificial intelligence (AI) algorithms, particularly deep learning, have demonstrated remarkable progress in image-recognition tasks. Methods ranging from convolutional neural networks to variational autoencoders have found myriad applications in the medical image analysis field, propelling it forward at a rapid pace. Historically, in radiology pr...

Radiomics aims to quantify phenotypic characteristics on medical imaging through the use of automated algorithms. Radiomic artificial intelligence (AI) technology, either based on engineered hard-coded algorithms or deep learning methods, can be used to develop noninvasive imaging-based biomarkers. However, lack of standardized algorithm definition...

Objective: Endoscopic endonasal approaches are increasingly performed for the surgical treatment of multiple skull base pathologies. Preventing postoperative CSF leaks remains a major challenge, particularly in extended approaches. In this study, the authors assessed the potential use of modern multimaterial 3D printing and neuronavigation to help...

We present a multimaterial voxel-printing method that enables the physical visualization of data sets commonly associated with scientific imaging. Leveraging voxel-based control of multimaterial three-dimensional (3D) printing, our method enables additive manufacturing of discontinuous data types such as point cloud data, curve and graph data, imag...

Radiographic imaging continues to be one of the most effective and clinically useful tools within oncology. Sophistication of artificial intelligence (AI) has allowed for detailed quantification of radiographic characteristics of tissues using predefined engineered algorithms or deep learning methods. Precedents in radiology as well as a wealth of...

Artificial intelligence (AI) algorithms hold the potential to revolutionize radiology. However, a significant portion of the published literature lacks transparency and reproducibility, which hampers sustained progress toward clinical translation. Although several reporting guidelines have been proposed, identifying practical means to address these...

Foundation models in deep learning are characterized by a single large-scale model trained on vast amounts of data serving as the foundation for various downstream tasks. Foundation models are generally trained using self-supervised learning and excel in reducing the demand for training samples in downstream applications. This is especially importa...

Foundation models represent a recent paradigm shift in deep learning, where a single large-scale model trained on vast amounts of data can serve as the foundation for various downstream tasks. Foundation models are generally trained using self-supervised learning and excel in reducing the demand for training samples in downstream applications. This...

Artificial intelligence (AI) algorithms hold the potential to revolutionize medical imaging. However, a significant portion of the published literature lacks transparency and reproducibility, which hampers sustained progress toward clinical translation. Although several reporting guidelines have been proposed, identifying practical means to address...

Background Artificial intelligence (AI) and deep learning have shown great potential in streamlining clinical tasks. However, most studies remain confined to in silico validation in small internal cohorts, without external validation or data on real-world clinical utility. We developed a strategy for the clinical validation of deep learning models...

Identifying the presence of intravenous contrast material on CT scans is an important component of data curation for medical imaging-based artificial intelligence model development and deployment. Use of intravenous contrast material is often poorly documented in imaging metadata, necessitating impractical manual annotation by clinician experts. Au...

Purpose/Objective(s) Automated tumor segmentation for oropharyngeal cancer (OPC) has the potential to improve treatment planning, response assessment, and clinical translation of imaging-based biomarkers. Deep learning has shown promise for cancer imaging segmentation, but performance for OPC tumors has been suboptimal with studies generally limite...

Purpose/Objective(s) Recent work has challenged emphasis of chronologic age in cancer treatment decision-making. For patients with non-small cell lung cancer (NSCLC), cardiopulmonary risk histories widely vary, possibly impacting “biologic age.” We aimed to apply an externally-derived deep learning model to estimate chest imaging age (ChIA) from di...

Purpose/Objective(s) Automated target segmentation for non-small cell lung cancer (NSCLC) patients has the potential to support radiation treatment planning. Artificial intelligence (AI) has demonstrated great promise in medical image segmentation tasks. However, most studies have been confined to in silico validation in small internal cohorts, lac...

Purpose: Identifying intravenous (IV) contrast use within CT scans is a key component of data curation for model development and testing. Currently, IV contrast is poorly documented in imaging metadata and necessitates manual correction and annotation by clinician experts, presenting a major barrier to imaging analyses and algorithm deployment. We...

Clinical oncology is experiencing rapid growth in data that are collected to enhance cancer care. With recent advances in the field of artificial intelligence (AI), there is now a computational basis to integrate and synthesize this growing body of multi-dimensional data, deduce patterns, and predict outcomes to improve shared patient and clinician...

Tumor histology is an important predictor of therapeutic response and outcomes in lung cancer. Tissue sampling for pathologist review is the most reliable method for histology classification, however, recent advances in deep learning for medical image analysis allude to the utility of radiologic data in further describing disease characteristics an...

Artificial intelligence (AI) has the potential to fundamentally alter the way medicine is practised. AI platforms excel in recognizing complex patterns in medical data and provide a quantitative, rather than purely qualitative, assessment of clinical conditions. Accordingly, AI could have particularly transformative applications in radiation oncolo...

In their study, McKinney et al. showed the high potential of artificial intelligence for breast cancer screening. However, the lack of detailed methods and computer code undermines its scientific value. We identify obstacles hindering transparent and reproducible AI research as faced by McKinney et al and provide solutions with implications for the...

Sharks and rays have distinctive skeletons among vertebrate animals, consisting primarily of unmineralized cartilage wrapped in a surface tessellation of minute polygonal tiles called tesserae, linked by unmineralized collagenous fibers. The discrete combination of hard and soft tissues is hypothesized to enhance the mechanical performance of tesse...

Significant efforts exist to develop living/non-living composite materials—known as biohybrids—that can support and control the functionality of biological agents. To enable the production of broadly applicable biohybrid materials, new tools are required to improve replicability, scalability, and control. Here, the Hybrid Living Material (HLM) fabr...

Man-made armors often rely on rigid structures for mechanical protection, which typically results in a trade-off with flexibility and maneuverability. Chitons, a group of marine mollusks, evolved scaled armors that address similar challenges. Many chiton species possess hundreds of small, mineralized scales arrayed on the soft girdle that surrounds...

Recent advances in artificial intelligence research have led to a profusion of studies that apply deep learning to problems in image analysis and natural language processing among others. Additionally, the availability of open-source computational frameworks has lowered the barriers to implementing state-of-the-art methods across multiple domains....

Artificial intelligence (AI) has demonstrated great progress in the detection, diagnosis, and treatment of diseases. Deep learning, a subset of machine learning based on artificial neural networks, has enabled applications with performance levels approaching those of trained professionals in tasks including the interpretation of medical images and...

Importance Chest radiography is the most common diagnostic imaging test in medicine and may also provide information about longevity and prognosis. Objective To develop and test a convolutional neural network (CNN) (named CXR-risk) to predict long-term mortality, including noncancer death, from chest radiographs. Design, Setting, and Participants...

Purpose: Tumors are continuously evolving biological systems, and medical imaging is uniquely positioned to monitor changes throughout treatment. Although qualitatively tracking lesions over space and time may be trivial, the development of clinically relevant, automated radiomics methods that incorporate serial imaging data is far more challengin...

Judgement, as one of the core tenets of medicine, relies upon the integration of multilayered data with nuanced decision making. Cancer offers a unique context for medical decisions given not only its variegated forms with evolution of disease but also the need to take into account the individual condition of patients, their ability to receive trea...

Background Non-small-cell lung cancer (NSCLC) patients often demonstrate varying clinical courses and outcomes, even within the same tumor stage. This study explores deep learning applications in medical imaging allowing for the automated quantification of radiographic characteristics and potentially improving patient stratification. Methods and f...

Stability against inter-reader variations. To simulate human readers annotating tumor centers with some variability, we translated the input seed point in 3D space. (A) Translation distances along X, Y, and Z are drawn separately from a binomial distribution with probabilities based on a normal distribution (σ = 4). Translations are limited to a 30...

Effects of tumor annotation information on prognostic power. The AUC plot illustrates the prognostic power of 3 different models as tested on the radiotherapy test dataset Maastro (n = 211). The first deep learning network, where the tumor volume is masked by giving regions beyond the tumor the value of air (−1,000 HU), is shown in green (AUC = 0.6...

Gene set enrichment analysis (GSEA) for MUMC dataset. (XLSX)

Benchmarking deep learning networks against engineered and clinical models. This figure compares the prognostic performance of deep learning networks with random forest models. The benchmarking is based on predicting overall 2-year survival. The deep learning networks are used for reference with performance at AUC = 0.70 (95% CI 0.63–0.78, p = 1.13...

Global gene set expression patterns—MUMC dataset. The deep learning network predictions on the surgery tuning dataset MUMC were linked to global gene expression patterns using a pre-ranked gene set enrichment analysis (GSEA). Negative and positive enrichments are shown in red and blue, respectively. The top 10 enrichments in each category are highl...

Benchmarking the effects of transfer learning—M-SPORE dataset. This plot illustrates the prognostic power of 3 different methodologies as tested on the surgery test dataset M-SPORE (n = 97). The first result (AUC = 0.71), shown in blue, represents the fine-tuned network with weights initialized from the radiotherapy network (Fig 3C). The second res...

Gene set enrichment analysis (GSEA) for Moffitt dataset. (XLSX)

Dataset breakdown. Table showing the 7 datasets used in this study: 3 radiotherapy datasets, 3 surgery datasets, and 1 stability assessment dataset. Only patients with NSCLC and stages I through III were selected. For Kaplan–Meier curves and genomic association studies, all patients with survival follow-up were used. For deep learning and engineere...

Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist. (DOCX)

Univariate Cox model results. Test results from univariate Cox models exploring the relationship between clinical factors and survival for both the radiotherapy and surgery patient groups. Deep learning was based on the median split from the respective tuning datasets. The same median split is used in the Kaplan–Meier curves (Fig 3B and 3D). (EPS)

Dataset information. (DOCX)

Distribution of tumor bounding box dimensions in the radiotherapy training dataset HarvardRT. This distribution is based on ground truth tumor annotations and was used to determine and optimize the input size to the CNN. An input size of 50 × 50 × 50 mm was found to be optimum as it gave the best performance on the tuning dataset Radboud. Around 60...

Background Successful transcatheter aortic valve replacement (TAVR) requires an understanding of how a prosthetic valve will interact with a patient's anatomy in advance of surgical deployment. To improve this understanding, we developed a benchtop workflow that allows for testing of physical interactions between prosthetic valves and patient-speci...

Objective: To design and validate a novel mixed reality head-mounted display for intraoperative surgical navigation. Design: A mixed reality navigation for laparoscopic surgery (MRNLS) system using a head mounted display (HMD) was developed to integrate the displays from a laparoscope, navigation system, and diagnostic imaging to provide context...

Radiographic imaging continues to be one of the most effective and clinically useful tools within oncology. Sophistication of artificial intelligence has allowed for detailed quantification of radiographic characteristics of tissues using predefined engineered algorithms or deep learning methods. Precedents in radiology as well as a wealth of resea...

Three-dimensional (3D) printing technologies are increasingly used to convert medical imaging studies into tangible (physical) models of individual patient anatomy, allowing physicians, scientists, and patients an unprecedented level of interaction with medical data. To date, virtually all 3D-printable medical data sets are created using traditiona...

Artificial intelligence (AI) systems built on incomplete or biased data will often exhibit problematic outcomes. Current methods of data analysis, particularly before model development, are costly and not standardized. The Dataset Nutrition Label (the Label) is a diagnostic framework that lowers the barrier to standardized data analysis by providin...

Objective: The purpose of this article is to describe a handheld external compression device used to facilitate CT fluoroscopy-guided percutaneous interventions in the abdomen. Conclusion: The device was designed with computer-aided design software to modify an existing gastrointestinal fluoroscopy compression device and was constructed by 3D pr...

The role of mixed reality that combines augmented and virtual reality in the healthcare industry, specifically in modern surgical interventions, has yet to be established. In laparoscopic surgeries, precision navigation with real-time feedback of distances from sensitive structures such as the pulmonary vessels is critical to preventing complicatio...

Tilings are constructs of repeated shapes covering a surface, common in both manmade and natural structures, but in particular are a defining characteristic of shark and ray skeletons. In these fishes, cartilaginous skeletal elements are wrapped in a surface tessellation, comprised of polygonal mineralized tiles linked by flexible joints, an arrang...

Voxelbeam explores precedents in the optimization of architectural structures, namely the Sydney Opera house Arup beam. The authors research three areas crucial to conceiving an innovative contemporary reinterpretation of the beam: A shift in structural analysis techniques from analytical to numerical models such as topology optimization, the funda...