Cost-effectiveness of Artificial Intelligence for Proximal Caries Detection.

Artificial intelligence (AI) can assist dentists in image assessment, for example, caries detection. The wider health and cost impact of employing AI for dental diagnostics has not yet been evaluated. We compared the cost-effectiveness of proximal caries detection on bitewing radiographs with versus without AI. U-Net, a fully convolutional neural network, had been trained, validated, and tested on 3,293, 252, and 141 bitewing radiographs, respectively, on which 4 experienced dentists had marked carious lesions (reference test). Lesions were stratified for initial lesions (E1/E2/D1, presumed noncavitated, receiving caries infiltration if detected) and advanced lesions (D2/D3, presumed cavitated, receiving restorative care if detected). A Markov model was used to simulate the consequences of true- and false-positive and true- and false-negative detections, as well as the subsequent decisions over the lifetime of patients. A German mixed-payers perspective was adopted. Our health outcome was tooth retention years. Costs were measured in 2020 euro. Monte-Carlo microsimulations and univariate and probabilistic sensitivity analyses were conducted. The incremental cost-effectiveness ratio (ICER) and the cost-effectiveness acceptability at different willingness-to-pay thresholds were quantified. AI showed an accuracy of 0.80; dentists' mean accuracy was significantly lower at 0.71 (minimum-maximum: 0.61-0.78, P < 0.05). AI was significantly more sensitive than dentists (0.75 vs. 0.36 [0.19-0.65]; P = 0.006), while its specificity was not significantly lower (0.83 vs. 0.91 [0.69-0.98]; P > 0.05). In the base-case scenario, AI was more effective (tooth retention for a mean 64 [2.5%-97.5%: 61-65] y) and less costly (298 [244-367] euro) than assessment without AI (62 [59-64] y; 322 [257-394] euro). The ICER was -13.9 euro/y (i.e., AI saved money at higher effectiveness). In the majority (>77%) of all cases, AI was less costly and more effective. Applying AI for caries detection is likely to be cost-effective, mainly as fewer lesions remain undetected. Notably, this cost-effectiveness requires dentists to manage detected early lesions nonrestoratively.

A synthetic lethal screen identifies SLK1, a novel protein kinase homolog implicated in yeast cell morphogenesis and cell growth.

The Saccharomyces cerevisiae SPA2 protein localizes at sites involved in polarized cell growth in budding cells and mating cells. spa2 mutants have defects in projection formation during mating but are healthy during vegetative growth. A synthetic lethal screen was devised to identify mutants that require the SPA2 gene for vegetative growth. One mutant, called slk-1 (for synthetic lethal kinase), has been characterized extensively. The SLK1 gene has been cloned, and sequence analysis predicts that the SLK1 protein is 1,478 amino acid residues in length. Approximately 300 amino acids at the carboxy terminus exhibit sequence similarity with the catalytic domains of protein kinases. Disruption mutations have been constructed in the SLK1 gene. slk1 null mutants cannot grow at 37 degrees C, but many cells can grow at 30, 24, and 17 degrees C. Dead slk1 mutant cells usually have aberrant cell morphologies, and many cells are very small, approximately one-half the diameter of wild-type cells. Surviving slk1 cells also exhibit morphogenic defects; these cells are impaired in their ability to form projections upon exposure to mating pheromones. During vegetative growth, a higher fraction of slk1 cells are unbudded compared with wild-type cells, and under nutrient limiting conditions, slk1 cells exhibit defects in cell cycle arrest. The different slk1 mutant defects are partially rescued by an extra copy of the SSD1/SRK1 gene. SSD1/SRK1 has been independently isolated as a suppressor of mutations in genes involved in growth control, sit4, pde2, bcy1, and ins1 (A. Sutton, D. Immanuel, and K.T. Arnat, Mol. Cell. Biol. 11:2133-2148, 1991; R.B. Wilson, A.A. Brenner, T.B. White, M.J. Engler, J.P. Gaughran, and K. Tatchell, Mol. Cell. Biol. 11:3369-3373, 1991). These data suggest that SLK1 plays a role in both cell morphogenesis and the control of cell growth. We speculate that SLK1 may be a regulatory link for these two cellular processes.

The KNS1 gene of Saccharomyces cerevisiae encodes a nonessential protein kinase homologue that is distantly related to members of the CDC28/cdc2 gene family.

A novel protein kinase homologue (KNS1) has been identified in Saccharomyces cerevisiae. KNS1 contains an open reading frame of 720 codons. The carboxy-terminal portion of the predicted protein sequence is similar to that of many other protein kinases, exhibiting 36% identity to the cdc2 gene product of Schizosaccharomyces pombe and 34% identity to the CDC28 gene product of S. cerevisiae. Deletion mutations were constructed in the KNS1 gene. kns1 mutants grow at the same rate as wild-type cells using several different carbon sources. They mate at normal efficiencies, and they sporulate successfully. No defects were found in entry into or exit from stationary phase. Thus, the KNS1 gene is not essential for cell growth and a variety of other cellular processes in yeast.

Studies concerning the temporal and genetic control of cell polarity in Saccharomyces cerevisiae.

The establishment of cell polarity was examined in the budding yeast, S. cerevisiae. The distribution of a polarized protein, the SPA2 protein, was followed throughout the yeast cell cycle using synchronized cells and cdc mutants. The SPA2 protein localizes to a patch at the presumptive bud site of G1 cells. Later it concentrates at the bud tip in budded cells. At cytokinesis, the SPA2 protein is at the neck between the mother and daughter cells. Analysis of unbudded haploid cells has suggested a series of events that occurs during G1. The SPA2 patch is established very early in G1, while the spindle pole body residues on the distal side of the nucleus. Later, microtubules emanating from the spindle pole body intersect the SPA2 crescent, and the nucleus probably rotates towards the SPA2 patch. By middle G1, most cells contain the SPB on the side of the nucleus proximal to the SPA2 patch, and a long extranuclear microtubule bundle intersects this patch. We suggest that a microtubule capture site exists in the SPA2 staining region that stabilizes the long microtubule bundle; this capture site may be responsible for rotation of the nucleus. Cells containing a polarized distribution of the SPA2 protein also possess a polarized distribution of actin spots in the same region, although the actin staining is much more diffuse. Moreover, cdc4 mutants, which form multiple buds at the restrictive temperature, exhibit simultaneous staining of the SPA2 protein and actin spots in a subset of the bud tips. spa2 mutants contain a polarized distribution of actin spots, and act1-1 and act1-2 mutants often contain a polarized distribution of the SPA2 protein suggesting that the SPA2 protein is not required for localization of the actin spots and the actin spots are not required for localization of the SPA2 protein. cdc24 mutants, which fail to form buds at the restrictive temperature, fail to exhibit polarized localization of the SPA2 protein and actin spots, indicating that the CDC24 protein is directly or indirectly responsible for controlling the polarity of these proteins. Based on the cell cycle distribution of the SPA2 protein, a "cytokinesis tag" model is proposed to explain the mechanism of the non-random positioning of bud sites in haploid yeast cells.

The SPA2 gene of Saccharomyces cerevisiae is important for pheromone-induced morphogenesis and efficient mating.

Upon exposure to mating pheromone, Saccharomyces cerevisiae undergoes cellular differentiation to form a morphologically distinct cell called a "shmoo". Double staining experiments revealed that both the SPA2 protein and actin localize to the shmoo tip which is the site of polarized cell growth. Actin concentrates as spots throughout the shmoo projection, while SPA2 localizes as a sharp patch at the shmoo tip. DNA sequence analysis of the SPA2 gene revealed an open reading frame 1,466 codons in length; the predicted protein sequence contains many internal repeats including a nine amino acid sequence that is imperfectly repeated 25 times. Portions of the SPA2 sequence exhibit a low-level similarity to proteins containing coiled-coil structures. Yeast cells containing a large deletion of the SPA2 gene are similar in growth rate to wild-type cells. However, spa2 mutant cells are impaired in their ability to form shmoos upon exposure to mating pheromone, and they do not mate efficiently with other spa2 mutant cells. Thus, we suggest that the SPA2 protein plays a critical role in cellular morphogenesis during mating, perhaps as a cytoskeletal protein.