[Computer-assisted skin cancer diagnosis : Is it time for artificial intelligence in clinical practice?] / Computerassistierte Hautkrebsdiagnose : Wann kommt künstliche Intelligenz in der Praxis an?

BACKGROUND: Artificial intelligence (AI) is increasingly being used in medical practice. Especially in the image-based diagnosis of skin cancer, AI shows great potential. However, there is a significant discrepancy between expectations and true relevance of AI in current dermatological practice. OBJECTIVES: This article summarizes promising study results of skin cancer diagnosis by computer-based diagnostic systems and discusses their significance for daily practice. We hereby focus on the analysis of dermoscopic images of pigmented and unpigmented skin lesions. MATERIALS AND METHODS: A selective literature search for recent relevant trials was conducted. The included studies used machine learning, and in particular "convolutional neural networks", which have been shown to be particularly effective for the classification of image data. RESULTS AND CONCLUSIONS: In numerous studies, computer algorithms were able to detect pigmented and nonpigmented neoplasms of the skin with high precision, comparable to that of dermatologists. The combination of the physician's assessment and AI showed the best results. Computer-based diagnostic systems are widely accepted among patients and physicians. However, they are still not applicable in daily practice, since computer-based diagnostic systems have only been tested in an experimental environment. In addition, many digital diagnostic criteria that help AI to classify skin lesions remain unclear. This lack of transparency still needs to be addressed. Moreover, clinical studies on the use of AI-based assistance systems are needed in order to prove its applicability in daily dermatologic practice.

Spontaneous mutation rates at five coat-color loci in mice.

Examination of 1.5 million mice yielded natural mutation rates estimnated from 5.2 million gene reproductions at five specific coat-color loci. The average rates were 11.1 x 10(-6) for forward mutations and 2.7 x 10(-6) for reverse mutations. Differences between the frequencies of mutations at the individual loci were evident.

Abnormality of calmodulin activity in hypertension. Evidence of the presence of an activator.

An apparent increase of calmodulin (CaM) activity was previously observed in the heart and kidney but not in the liver of spontaneously-hypertensive rats (SHR) and mice compared with their corresponding normotensive controls. As this change was due to an elevated recovery of CaM in the organs of the hypertensive animals, the present study was designed to evaluate its activity in hypertension. A CaM activator, detected in heart and kidney supernatants from hypertensive animals, was found to be responsible for this enhanced recovery. Similar results were obtained with passaged, cultured aortic smooth muscle cells from SHR, indicating that the anomaly was not a mere consequence of elevated blood pressure but rather a genetic expression of cells of hypertensive origin. The activator was heat stable, nondialyzable, and recovered in the fraction precipitated with 30-50% ammonium sulfate. Preliminary extraction studies suggest that the activator is contained in a glycolipid fraction. The stimulation of phosphodiesterase by this activator was calcium and CaM dependent. The activator appears to affect the affinity of the phosphodiesterase for CaM rather than the maximal stimulation. The activator was also present at a low concentration in the heart and kidney of normotensive animals. These findings indicate that at least some of the calcium abnormalities implicated in the pathogenesis of hypertension could be the result of interactions between CaM, calcium, and this activator.

Chronic hypertension and altered baroreflex responses in transgenic mice containing the human renin and human angiotensinogen genes.

We have generated a transgenic model consisting of both the human renin and human angiotensinogen genes to study further the role played by the renin-angiotensin system in regulating arterial pressure. Transgenic mice containing either gene alone were normotensive, whereas mice containing both genes were chronically hypertensive. Plasma renin activity and plasma angiotensin II levels were both markedly elevated in the double transgenic mice compared with either single transgenic or nontransgenic controls. The elevation in blood pressure caused by the human transgenes was independent of the genotype at the endogenous renin locus and was equal in mice homozygous for the Ren-1c allele or in mice containing one copy each of Ren-1c, Ren-1d, or Ren-2. Chronic overproduction of angiotensin II in the double transgenic mice resulted in a resetting of the baroreflex control of heart rate to a higher pressure without significantly changing the gain or sensitivity of the reflex. Moreover, this change was not due to the effects of elevated pressure itself since angiotensin-converting enzyme inhibition had minimal effects on the baroreflex in spontaneously hypertensive BPH-2 control mice, which exhibit non-renin-dependent hypertension. This double transgenic model should provide an excellent tool for further studies on the mechanisms of hypertension initiated by the renin-angiotensin system.

Selection for blood pressure levels in mice.

Response to two-way selection for systolic blood pressure was immediate and continuous for about eight generations. In the twelfth generation, the High males differed from the Low males by 38 mmHG; the females differed by 39 mmHg. There was little overlap between the two lines and they were statistically significant from each other and from the Random control line. There appeared to be no more additive genetic variance in the eleventh and twelfth generations. Causes for the cessation of response are explored. This is probably due to a combination of natural selection acting to reduce litter sizes in the Low line, a higher incidence of sudden deaths in the High line, and loss of favorable alleles as both selection lines went through a population bottleneck in the ninth generation.-In the eleventh generation, the selected lines were used to produce F(1), F(2), and backcross generations. A genetic analysis yielded significant additive and dominance components in the inheritance of systolic blood pressure.

Thermosensitivity, a possible new locus involved in genetic hypertension.

Spontaneously hypertensive mice have been characterized as more sensitive to environmental heat than normotensive mice. A breeding program was therefore initiated to examine the possible genetic link between thermosensitivity and hypertension. Crossbreeding of spontaneously hypertensive mice with randomly bred normotensive mice produced F1 hybrids, which were then intercrossed to create a F2 population. Thermosensitivity was measured with a noninvasive method. The rate of body temperature increase was significantly (p less than 0.001) higher in the hypertensive mice (1.74 +/- 0.04 degrees C/min) compared with normal controls (1.13 +/- 0.03 degrees C/min). The frequency distribution of the rate of body temperature increase among the progenies was consistent with the hypothesis that a single gene locus determines the observed difference in thermosensitivity between normal and hypertensive mice. The allele that determines the rate of body temperature increase in normal mice was dominant in relation to the allele contributed by hypertensive mice. In the F2 population, a bimodal distribution determined two phenotypes: less than 1.40 degrees C/min and greater than 1.40 degrees C/min. A significant difference (p less than 0.01) in blood pressure of 11 mm Hg was observed between these two phenotypes. In addition, a positive correlation (p less than 0.01) was noted between the rate of body temperature increase and blood pressure in the F2 progeny. We conclude that there is possibly a single locus controlling thermosensitivity, which exhibits additive-dominance inheritance. Alleles of this particular trait segregate in part with an increment in blood pressure. The results support the possibility that the increased thermosensitivity seen in hypertensive mice is associated with one of the genes that contributes to their high blood pressure.

Genome scan for blood pressure loci in mice.

Hypertension is a complex trait of unknown cause in humans. Mice of the inbred strain BPH/2 serve as a rodent model of human hypertension and display elevated blood pressure compared with the hypotensive strain BPL/1. An F2 intercross of BPH/2 and BPL/1 and 2 backcrosses of BPL/1 with Mus spretus were used to perform interval linkage mapping for systolic blood pressure in a genome scan. Significant linkage was observed in the F2s on chromosome 10 (logarithm of the odds score [LOD]=4.9) and on chromosome 13 in the M spretus backcross (LOD=3.3), with additional suggestive LODs on chromosomes 2, 6, 8, and 18. In addition, several suggestive linkages were observed for phenotypes associated with human hypertension. Our study is the first reported genome-wide linkage scan for blood pressure genes in the mouse.

Decrease of blood pressure in spontaneously hypertensive mice by heat treatment.

Although the increased sensitivity of hypertensive animals to heat stress has been reported, the effect of chronic heat exposure has not been examined. The specific goal of the present investigation was to study the impact of chronic heat treatment on the blood pressure of spontaneously hypertensive mice. Chronic 40 degrees C heat exposure for 5 min daily progressively lowered basal blood pressure in hypertensive mice within 20 days, without any change in normal mice. In fact, after 35 days of chronic heat treatment, the basal blood pressure of hypertensive mice was indistinguishable from that of the normotensives. Repeated immobilization and prewarming as normal procedures for recording blood pressure contributed to the decrease in blood pressure by 10 to 12 mm Hg, but chronic heat by itself was significantly more potent in reducing it by an additional 20 mm Hg. After the discontinuation of chronic heat application, the basal blood pressure of hypertensive mice returned with time to the level registered in sham-handled hypertensive controls. These results demonstrate that, although acute heat is more detrimental to hypertensive mice, brief, chronic exposure to mild heat stress is beneficial in that it normalizes basal blood pressure.

Catecholamine concentrations in discrete brain nuclei and sympathetic tissues of genetically hypertensive mice.

Catecholamine concentrations were determined at central and peripheral level in high blood pressure (HBP), low blood pressure (LBP) and random bred (RB) mice. In HBP mice compared to LBP, the noradrenaline concentration was lower in the locus coeruleus, medullary A1-C1 and A2-C2 areas, thoracic spinal cord, posteroventral hypothalamus and nucleus hypothalamicus anterior, while dopamine concentration was decreased in the striatum. Adrenal catecholamine levels were higher in both HBP and LBP compared to RB mice.

Consumption of electrolytes and quinine by mouse strains with different blood pressures.

Daily fluid intakes were measured using two-bottle tests in female mice of inbred strains with high (BPH/2), normal (BPN/3) or low (BPL/1) blood pressure. The mice were offered a choice between water and different concentrations of NaCl (37.5-600 mM), KCl (1-400 mM), CaCl2 (1-100 mM) and quinine hydrochloride (0.003-1.0 mM). Compared with the normotensive strain, the hypertensive mice had higher water and total fluid intakes, and lower intakes of NaCl, KCl (only 200 mM) and quinine; the hypotensive mice had higher intakes of KCl (only 10-50 mM) and lower intakes of CaCl2 and quinine. These data suggest that fluid and salt intake are not linearly related to blood pressure, but are independently determined in these strains. Certain concentrations of the salts were preferred relative to water, which depended on mouse genotype: the BPN/3 and BPL/1 mice strongly preferred 37.5-150 mM NaCl, the BPL/1 mice preferred 10-100 mM KCl, and the BPN/3 mice preferred 1-10 mM CaCl2.

[Central and peripheral catecholamines in the genetically hypertensive mouse]. / Catécholamines centrales et périphériques chez la souris génétiquement hypertendue.

The aim of this work was to assess if biochemical alterations are present in catecholaminergic neurons of genetically hypertensive mice (HBP strain) compared to genetically hypotensive mice (LBP strain), as well as to mice of the population (R) from which HBP and LBP strains have been selected. For that purpose, dopamine (DA), noradrenaline (NA) and adrenaline (A) concentrations were determined by high performance liquid chromatography with electrochemical detection, in peripheral structures and brain nuclei of adult male mice from these three strains. In the adrenal medulla of HBP mice, DA, NA and A concentrations are increased when compared to R, but not to LBP mice. In the superior cervical ganglia of HBP mice, the NA concentration is decreased when compared to LBP, but increased when compared to R mice. At the central level, the NA concentration is decreased in HBP mice when compared to LBP, but not significantly changed when compared to R mice, in the following regions: locus coeruleus (-39 p. 100, p less than 0.001), A1-C1 (-17 p. 100, p less than 0.001) and A2-C2 (-19 p. 100, p less than 0.001) areas of the medulla oblongata, thoracic spinal cord (-26 p. 100, p less than 0.01), posteroventral hypothalamus (-18 p. 100, p less than 0.01) and nucleus hypothalamicus anterior (-14 p. 100, p less than 0.01). In the nucleus periventricularis, NA concentration of HBP mice is not changed when compared to both LBP and R animals. Finally, in the caudate nucleus of HBP mice, the DA concentration is decreased when compared to both R and LBP mice.(ABSTRACT TRUNCATED AT 250 WORDS)

Prevention of neonatal oxygen-induced brain damage by reduction of intrinsic apoptosis.

Within the last decade, it became clear that oxygen contributes to the pathogenesis of neonatal brain damage, leading to neurocognitive impairment of prematurely born infants in later life. Recently, we have identified a critical role for receptor-mediated neuronal apoptosis in the immature rodent brain. However, the contribution of the intrinsic apoptotic pathway accompanied by activation of caspase-2 under hyperoxic conditions in the neonatal brain still remains elusive. Inhibition of caspases appears a promising strategy for neuroprotection. In order to assess the influence of specific caspases on the developing brain, we applied a recently developed pentapeptide-based group II caspase inhibitor (5-(2,6-difluoro-phenoxy)-3(R,S)-(2(S)-(2(S)-(3-methoxycarbonyl-2(S)-(3-methyl-2(S)-((quinoline-2-carbonyl)-amino)-butyrylamino)propionylamino)3-methylbutyrylamino)propionylamino)-4-oxo-pentanoic acid methyl ester; TRP601). Here, we report that elevated oxygen (hyperoxia) triggers a marked increase in active caspase-2 expression, resulting in an initiation of the intrinsic apoptotic pathway with upregulation of key proteins, namely, cytochrome c, apoptosis protease-activating factor-1, and the caspase-independent protein apoptosis-inducing factor, whereas BH3-interacting domain death agonist and the anti-apoptotic protein B-cell lymphoma-2 are downregulated. These results coincide with an upregulation of caspase-3 activity and marked neurodegeneration. However, single treatment with TRP601 at the beginning of hyperoxia reversed the detrimental effects in this model. Hyperoxia-mediated neurodegeneration is supported by intrinsic apoptosis, suggesting that the development of highly selective caspase inhibitors will represent a potential useful therapeutic strategy in prematurely born infants.

Systemic G-CSF treatment does not improve long-term outcomes after neonatal hypoxic-ischaemic brain injury.

Hypoxia-ischaemia (HI) is a major factor in the pathogenesis of developmental brain injury, leading to cognitive deficits and motor disabilities in preterm infants. The haematopoietic growth factor granulocyte colony-stimulating factor (G-CSF) has been shown to exert a neuroprotective activity in rodent models of ischaemic stroke and is currently subject to phase I/II clinical trials in adults. Results of studies examining the effect of G-CSF in perinatal brain damage have been contradictory. We have previously shown that G-CSF increases NMDAR-mediated excitotoxic brain injury in the neonatal mouse brain. In this study, we evaluated the effect of G-CSF on long-term outcomes after HI. On postnatal day 5, mice pubs were first randomly assigned to a sham operation or HI and then divided into four treatment groups: i) G-CSF; ii) phosphate buffered saline (PBS) 1h after injury; iii) G-CSF and iv) PBS 60 h after injury. G-CSF (200 µg/kg BW) was administered five times within a 24h interval. Neuromotor and cognitive outcomes were assessed by open-field, novel object recognition tests and rotarod tests starting on P90, with subsequent histological analyses of brain injury. G-CSF treatment did not improve either neurobehavioural outcomes or brain injuries. Interestingly, the application of PBS and G-CSF in the acute phase increased brain damage in the hippocampus. We could not confirm the neuroprotective properties of G-CSF in neonatal HI brain damage. The exacerbation of injury by the administration of substances in the acute phase might indicate a heightened state of neurological sensitivity that is specific to mechanisms of secondary neurodegeneration and influenced by unidentified external factors possibly associated with the treatment protocol during the acute phase. This article is part of a Special Issue entitled "Interaction between repair, disease, & inflammation."