Like most connectionist systems, deep networks have been found to be prone to catastrophic forgetting effects. This makes generalization of deep neural network pipelines a challenge as new additions to prediction requirements at runtime would invariably require retraining on not only the new dataset, but also substantial portions of older task data. This is a difficult task in personalised clinical imaging where retention of datasets over extended time is challenged by the fact that the already limited individual-specific examples may not be stored over extended time owing to memory, legal and infrastructure constraints. Thus, there is a need to rethink the process of deploying learning pipelines in personalised healthcare contexts that address forgetting as part of initial and incremental task learning over time. This has been modelled as an incremental learning problem, with an evolving interest in exploring the applicability of such a paradigm in precision medicine contexts. We propose a novel approach to the incremental class addition problem for individualised data curation settings pertinent to personalised medicine, where a retention of limited numbers of exemplars of old classes helps reduce forgetting instead of large-scale data storage, using a strategy of incremental time sample augmentation with fractional linear transformations and weighted knowledge distillation objectives to correct for evolving class imbalance effects in conjunction with the mitigation of reductions in initial task performances following incremental adaptation to new data distributions.