The implementation of clinical-decision support algorithms for medical imaging faces challenges with reliability and interpretability. Here, we establish a diagnostic tool based on a deep-learning framework for the screening of patients with common treatable blinding retinal diseases. Our framework utilizes transfer learning, which trains a neural network with a fraction of the data of conventional approaches. Applying this approach to a dataset of optical coherence tomography images, we demonstrate performance comparable to that of human experts in classifying age-related macular degeneration and diabetic macular edema. We also provide a more transparent and interpretable diagnosis by highlighting the regions recognized by the neural network. We further demonstrate the general applicability of our AI system for diagnosis of pediatric pneumonia using chest X-ray images. This tool may ultimately aid in expediting the diagnosis and referral of these treatable conditions, thereby facilitating earlier treatment, resulting in improved clinical outcomes. VIDEO ABSTRACT.

No abstract

Zuk et al in 2001 identified stem and regenerative cells within the stromal vascular fraction of fat. In preclinical studies, these cells appeared to stimulate angiogenesis and reduce inflammation, and soon thereafter, clinical use of stromal vascular fraction (SVF) evolved as researchers such as Rigotti, Coleman, Mojallal, our group, and others demonstrated that fat can be used for both therapeutic and aesthetic indications. The regenerative effects of fat and its contents on facial aesthetics have been shown at the histologic and cellular level. Regeneration of elastin and collagen fibers as well as improvement in capillary density and reduction of inflammation have been reported. We review our current approach to the use of regenerative cells and different types of fat grafts in facial surgery. The fat graft is classified, both from a regenerative point of view as well as a tissue product that can be modified into different tissue characteristics, depending on the area and condition treated. Clinical use of SVF enriched fat, millifat, microfat, and nanofat grafts as well as composite fat grafts are reviewed. Based on clinical experience and evidence to date, it appears that the regenerative effects seen with the use of SVF in aesthetic surgery are modest, but there appear to be definite histologic findings of regeneration. These improvements may not be clinically apparent to a patient when cell enriched fat grafts are compared to fat grafts alone. However, the subtle changes seen in histology may be cumulative over time. Three types of fat grafts are defined: millifat (parcel size 2.4<), microfat (1.2<), and nanofat (400-600 μm). Each are characterized by their injectability ratings and emulsification parcel size as well as amount of sSVF cells. Newer concepts of periosteal fat grafting, buccal fat pad grafting, pyriform aperture fat grafting, intraorbital fat grafting, and nanofat grafting are discussed. Composite fat grafts are presented as a new concept as is biofilling and biocontouring. The use of regenerative cells in facial surgery is evolving rapidly. Our understanding of the anatomic changes that occur with aging has become more precise and our ability to target histologic changes seen with aging has become more effective. Deep fat compartment grafting, superficial fat grafting, nanofat, and SVF are becoming important components of contemporary facial rejuvenation. The use of regenerative approaches in facial rejuvenation is a logical step in changing the paradigm from surgical treatment of aging to a more proactive prevention and maintenance approach that keeps up with changes in the tissues as they age.

Background Enzymatic digestion has been the gold standard for stromal vascular fraction (SVF) isolation but remains expensive while holding practical and legal concerns. Mechanical SVF isolation methods have been known to produce lower cell yields, but may potentially produce a more robust product by preserving the extracellular matrix (ECM) niche. Objectives Compare the relative efficacy on wound healing between mechanically-dissociated SVF (M-SVF) and enzymatically-digested SVF (E-SVF) Methods Lipoaspirate was partitioned into two equal groups and processed by either mechanical or enzymatic isolation methods. After SVF isolation, cell counts and viabilities were determined by flow cytometry and cell proliferation rates were measured by WST-1 test. Wound healing Scratch Assay test, which is commonly used to model in-vitro wound healing, was performed on both cell cocktails. Collagen type 1 (Col1A) gene expression level, which is known for its role in wound healing, was also measured for both groups. Results As predicted, E-SVF isolated more cells (1.74x106/ml [±3.63, n=10, p=0.015]) than M-SVF (0.94x106/ml [±1.69, n=10, p=0.015]), but no significant difference was observed in cell viability. However, M-SVF expressed over 2-fold higher stem cell surface markers and a 10% higher proliferation rate compared to E-SVF. In addition, the migration rate and level of Col1A gene expression of M-SVF were found to be significantly higher than E-SVF. Conclusion Although the cell yield of M-SVF was less than the E-SVF, M-SVF appears to have superior wound healing properties.